Volume 13 • Number 1 • Winter 2001 Journal of the Urban and Regional …€¦ · · 2016-05-12Journal of the Urban and Regional Information Systems Association CONTENTS

Volume 13 • Number 1 • Winter 2001

Journal of the Urban and Regional Information Systems Association

CONTENTS

REFEREED

5 The Concept and Implementation of Perceptual Regions as Hierarchical SpatialUnits for Evaluating Environmental SensitivityIsabelle Reginster and Geoffrey Edwards

17 Implementing the Enterprise GIS in TransportationDatabase DesignJ. Allison (Al) Butler and Kenneth J. Dueker

29 Constructing GIS: Actor Networks of CollaborationFrancis Harvey

PLUS

39 Review of Current Journal Literature

47 Information Technology Reports and Studies: Update

On the CoverTransportation dominates our daily lives from the morning commute, to thesummer vacation, to the importance of getting an 18-wheeler to its destina-tion. The use of enterprise GIS within the context of transportation is revo-lutionizing the way people and products crisscross the country. Our highwaysand bi-ways are the veins that pump the lifeblood of commerce, industry,and personal freedom to this nation and others. Implementing enterpriseGIS in transportation database design is the subject of an article by J. AllisonButler, the GIS Director for Hamilton County, Tennessee and KennethDueker, who is the Director of Transportation Studies at Portland State Uni-versity. Practical models, examples, and solutions for dealing with transpor-tation related issues are presented in their paper, which highlights the Winterissue of the URISA Journal.

2 URISA Journal • Vol. 12, No. 4 • Winter 2001

Journal

EDITORIAL OFFICE: Urban and Regional Information Systems Association, 1460 RenaissanceDrive, Suite 305, Park Ridge, Illinois 60068-1348; Voice (847) 824-6300; Fax (847) 824-6363;E-mail [email protected].

SUBMISSIONS: This publication accepts from authors an exclusive right of first publication totheir article plus an accompanying grant of non-exclusive full rights. The publisher requires that fullcredit for first publication in the URISA Journal is provided in any subsequent electronic or printpublications. For more information, the “Manuscript Submission Guidelines for Refereed Articles”is available on our web site, www.urisa.org, or by calling (847) 824-6300.

SUBSCRIPTION AND ADVERTISING: All correspondence about advertising, subscriptions,and URISA memberships should be directed to: Urban and Regional Information Systems Associa-tion, 1460 Renaissance Dr., Suite 305, Park Ridge, Illinois, 60068-1348; Voice (847) 824-6300;Fax (847) 824-6363; E-mail [email protected].

URISA Journal is published four times a year by the Urban and Regional Information SystemsAssociation.

© 2000 by the Urban and Regional Information Systems Association. Authorization to photocopyitems for internal or personal use, or the internal or personal use of specific clients, is granted bypermission of the Urban and Regional Information Systems Association.

Educational programs planned and presented by URISA provide attendees with relevant and rewarding continuing education experience. However,neither the content (whether written or oral) of any course, seminar, or other presentation, nor the use of a specific product in conjunctiontherewith, nor the exhibition of any materials by any party coincident with the educational event, should be construed as indicating endorsement orapproval of the views presented, the products used, or the materials exhibited by URISA, or by its committees, Special Interest Groups, Chapters, orother commissions.

SUBSCRIPTION RATE: One year: $295 business, libraries, government agencies, and public institutions. Individuals interested in subscriptionsshould contact URISA for membership information.

US ISSN 1045-8077

Publisher: Urban and Regional Information Systems Association

Editor-in-Chief: Harlan Onsrud

Journal Coordinator: Scott A. Grams

Interm Executive Directors: Wendy FrancisAnna Mae Scolaro

Electronic Journal: http://www.urisa.org/journal.htm

Correction:Within the Fall Issue of the URISAJournal, Vol 12, No. 4, an incor-rect reference was listed in an ar-ticle by Shilpam Pandey et. al.entitled Developing a Web-enabledTool to Assess Long Term HydrologicImpacts of Land-use Change: Infor-mation Technology Issues and a CaseStudy. The reference on page 21should read Klein, D.H., 1999,The Internet and the Future of theDecision Making Process. URISA1998 Annual Conference Proceed-ings Charlotte, North Carolina,August 1998. URISA regrets anyinconvenience.

URISA Journal • Vol. 13, No. 1 • Winter 2001 3

URISA Journal EditorEditor-in-Chief

Harlan Onsrud, Spatial Information Science andEngineering, University of Maine

Thematic EditorsEditor-Urban and Regional InformationScience

Lewis Hopkins, Department of Planning,University of Illinois-Champaign/Urbana

Editor-Applications ResearchLyna Wiggins, Department of Planning,Rutgers University

Editor-Social, Organizational, Legal, andEconomic Sciences

Ian Masser, Department of Urban Planningand Management, ITC (Netherlands)

Editor-Geographic Information ScienceMichael Goodchild, Department of Geography,University of California-Santa Barbara

Editor-Information and Media SciencesMichael Shiffer, Department of Planning,Massachusetts Institute of Technology

Editor-Spatial Data Acquisition and IntegrationGary Hunter, Department of Geomatics,University of Melbourne (Australia)

Editor-Geography, Cartography, andCognitive Science

David Mark, Department of Geography,SUNY-Buffalo

Editor-EducationKaren Kemp, Department of Geography,University of California-Berkeley

Section EditorsSoftware Review Editor

Jay Lee, Geography, Kent State University

Book Review EditorRebecca Somers, Somers-St. Clair

Literature Review Editor Zorica Nedovic, University of Illinois-Champaign/Urbana

Article Review BoardPeggy Agouris, Department of Spatial Informa-tion Science and Engineering, University of Maine

Michael Batty, Centre for Advanced Spatial Analy-sis, University College London (United Kingdom)

Kate Beard, Department of SpatialInformation Science and Engineering,University of Maine

Yvan Bédard, Centre for Research inGeomatics, Laval University (Canada)

Barbara P. Buttenfield, Department of Geog-raphy, University of Colorado

Keith C. Clarke, Department of Geography,University of California-Santa Barbara

David Coleman, Department of Geodesy andGeomatics Engineering, University of NewBrunswick (Canada)

David J. Cowen, Department of Geography,University of South Carolina

Massimo Craglia, Department of Town &Regional Planning, University of Sheffield(United Kingdom)

William J. Craig, Center for Urban and Re-gional Affairs, University of Minnesota

Robert G. Cromley, Department of Geogra-phy, University of Connecticut

Kenneth J. Dueker, Urban Studies andPlanning, Portland State University

Geoffrey Dutton, Spatial Effects

Max J. Egenhofer, Department of Spatial Informa-tion Science and Engineering, University of Maine

Manfred Ehlers, Geoinformatics and Institutefor Environmental Sciences, University ofVechta (Germany)

Manfred M. Fischer, Economics, Geography& Geoinformatics, Vienna University of Eco-nomics and Business Administration (Austria)

Myke Gluck, School of Information Studiesand Geography, Florida State University

Michael Gould, Department of Science,Experimentales Universitat (Spain)

Daniel A. Griffith, Department of Geography,Syracuse University

Francis J. Harvey, Department of Geography,University of Kentucky

Kingsley E. Haynes, Public Policy and Geog-raphy, George Mason University

Eric J. Heikkila, School of Policy, Planning, andDevelopment, University of Southern California

Stephen C. Hirtle, Department of Infor-mation Science and Telecommunications,University of Pittsburgh

Richard E. Klosterman, Department of Geog-raphy and Planning, University of Akron

Robert Laurini, Claude Bernard Universityof Lyon (France)

Thomas M. Lillesand, EnvironmentalRemote Sensing Center, University of Wisconsin

Xavier R. Lopez, Oracle Corporation

David Maguire, Environmental Systems Re-search Institute

John McLaughlin, Research and Interna-tional Cooperation, University of NewBrunswick (Canada)

Harvey J. Miller, Department of Geography,University of Utah

Joel L. Morrison, Center for Mapping, OhioState University

Atsuyuki Okabe, Department of Urban En-gineering, University of Tokyo (Japan)

Jeffrey K. Pinto, School of Business, PennState Erie

Gerard Rushton, Department of Geography,University of Iowa

Bruce D. Spear, Geographic InformationServices Bureau of Transportation Statistics,Washington, D.C.

Jonathan Sperling, Geography Division, U.S.Census Bureau

David J. Unwin, School of Geography,Birkbeck College, London (United Kingdom)

Stephen J. Ventura, Environmental Studies andSoil Science, University of Wisconsin-Madison

Nancy von Meyer, Fairview Industries

Barry Wellar, Department of Geography,University of Ottawa (Canada)

Michael F. Worboys, Department of ComputerScience, Keele University (United Kingdom)

EDITORS AND REVIEW BOARD

Model JobDescriptions forGIS ProfessionalsEdited by WilliamHuxhold

Innovative technology,increased demand, andnewfound applicationshave caused the field ofGIS to expand. As thisfield became morecomplex and varied,

the individual job requirements of the GIS professional beganto splinter. This new publication provides a collection of jobdescriptions you can use and modify to meet the needs ofyour own organization.

Based on URISA’s recent research on GIS job classifications instate and local government, the model job descriptions aredivided into six special categories based on job responsibility.GIS managers, coordinators, systems analysts/programmers,data analysts, specialists, and technicians are broken downand analyzed according to educational background, salaryrange, specific skills, personnel supervised, etc. Most commonjob titles and responsibilities are provided in each of the sixcategories.

This is an extremely helpful publication to anyone just settingup a GIS, professionals between careers, human resourcemanagers and to those who are seeking the most currentinformation on the state and identity of the GIS community.The GIS job classification areas correspond to the new URISA2000 Salary Survey data available in a separate publicationfrom URISA.

Member price $49, Nonmember price $69

Whether you are a manager who must hire and develop IT & GISstaff or just want to explore your own options...get ready for thenew year with these helpful publications!

URISA 2000Salary Survey forIT/GISProfessionalsEdited by PeterCroswell andNina Savar

1998 marked thefirst URISA SalarySurvey. Based on itssuccess and thehigh-level of

demand for more information by URISA members, the 2000Salary Survey has been greatly expanded and includes a muchwider-range of detailed information. Additions includenumber of staff members, computer skill requirements,educational level, years of experience, and department inwhich the GIS or IT professional works.

This new publication provides the answers to these importantquestions:� What are the differences in pay between GIS and/or IT

managers, analysts, coordinators?� What additional forms of compensation/benefits do

organizations offer?� How does your pay and benefits package stack up against

your peers?� Where are the highest paying jobs?� What are the most popular programming languages,

operating systems and software packages used in GISoffices?

� How much do you need in your budget in order to attractand hire IT and GIS professionals at all levels?

The 2000 URISA Salary Survey has this information andmuch more. The data presented has been culled from theresults of URISA’s salary survey that targeted a comprehensiveand representative sample of both the IT and GIScommunities. Salary data is presented according to region, jobtitle, type of organization, etc.

This publication also features an updated and expanded freebonus article to help you find, hire and keep technologyprofessionals. This publication is vital to anyone who hires IT/GIS professionals or to those pursuing careers in the field.Place your order now to be sure you receive a copy as soon as itis available. Shipping begins December, 2000.

Member price $49, Nonmember price $69

visit www.urisa.org/store.htm

or call

URISA Headquarters at

(847) 824-6300

UR

ISA C

are

er

Cente

r Public

ations

To order

URISA Journal � Reginster, Edwards 5

The Concept and Implementation of Perceptual Regions asHierarchical Spatial Units for Evaluating Environmental

Sensitivity

Isabelle Reginster and Geoffrey Edwards

Abstract: Studies of human spatial behavior and life spaces are useful because they allow a better understanding of the relation-ship between people and environment. In an era when there is growing public pressure to understand this relationship, studies oflife spaces may provide insight into the environmental sensitivity of different groups of people. In this article, we propose amethod for characterizing life spaces based on perceptual factors and show how the method can be used to explore the sensitivityof humans to environmental quality and to study human spatial behavior in the form of residential choice. In particular, theperceptual regions that we introduce constitute new hierarchical spatial units of analysis that join location to activities, the twokey concepts of life spaces. The structural and environmental differences of the perceptual regions in relation to the sociologicalcharacteristics of urban and suburban behaviors are explored for two districts within the city of Québec (Canada). The approachoffers the potential for developing some interesting applications in urban planning: the means to be more sensitive to the wishesof individual households in decisions concerning urban spaces, and a tool to assist persons evaluating different residential loca-tions. It is noted that the analyses performed may be largely automated.

IntroductionHuman spatial behavior expresses itself in many ways (Golledgeand Stimson 1997). One of the most interesting, albeit highlycomplex, forms of such expression is the creation of life spaces,particularly built environments, such as are found in cities.

The concepts and methodologies currently used for study-ing life spaces are far from being widely accepted. The variety ofapproaches proposed is a result of both the complexity and themultidisciplinary nature of the subject. The best example of thiscomplexity is probably the great variety of terms that are used todefine life spaces. These vary according to both the disciplinesconcerned and the spatial scale. At its origin, the term “humanterritoriality” is a biological term. It concerns the locations thatpeople own, occupy, or use for varying periods of time (Taylor1988). The notion of “place,” on the other hand, taken fromgeography, represents both a position in society and a spatial lo-cation (Tuan 1996). Within this view, places are foci where weexperience the meaningful events of our existence and points ofdeparture from which we orient ourselves and take possession ofthe environment (Norberg and Schultz 1971). “Spatial familiar-ity” is a concept that links space and cognition. It implies a stateof knowledge brought about by repeated association with thespace concerned (Gale et al. 1990). “Neighborhood” is the mostgeneral term and has been defined as a series of dimensions thatare independent of each other in certain situations while being

highly interrelated in other situations (Lyon 1987).It appears from these definitions that the study of life spaces

depends mainly on where people are and what they do. The twokey concepts in each of these analyses are location and activity(Garling and Golledge 1993).

Research on the Implementation of HumanSpaces in AnalysesAmong the many recent approaches to the definition of life spacesand spatial perception, several address recommendations, diffi-culties, and perspectives for implementations. Crevoisier (1996)highlights three particular difficulties in the formalization of theconcept of territory: the consideration of both physical and socio-economic proximities, the superimposition of the dimensions ofsocial life, and the specificity of the territory in each case.

An important concept in the definition of human spaces isto consider the hierarchical interrelation of spaces. The evidenceof hierarchical processing is emphasized by Fotheringham andCurtis (1992). Human sensitivity to space is at the center of agreat deal of research, but the implementation or structuring ofthese hierarchies is often different from one work to the next.Among others, Hirtle and Jonides (1985) and Hirtle (1995) de-velop hierarchies based on the use of trees and lattices that werein turn adapted by McNamara (1986). On the other hand, Davis(1981) introduced inferences between cross-hierarchical repre-


sentations. The conceptual task of modeling hierarchies is verydifficult, perhaps because of the complexity of human reasoning.

The means for incorporating the concept of the “home” inanalyses have been extensively explored. The meaning of houseand home in the United States was reviewed by John Adams(1984). Waddell and Shukla (1993) emphasized the importanceof elaborating theories about home in a variety of local settings.Shanton (1986) elaborated on different types of areal conceptassociated with the idea of “home.” Aitken and Prosser (1990)further studied these areal forms, based on cognitive and behav-ioral data. Their findings indicate that there is a structural differ-ence in the spatial familiarity of residents who perceive theirneighborhood as an area and those who perceive it as a network.

Edwards (1997, 2000) developed a simulation of humanspatial memory that structures experiential space based on rein-forcement (see discussion of spatial familiarity above). This modeldoes not distinguish in any fundamental way between externallyexperienced events and internally experienced events. Only ac-tions are given distinct status. As a result, internal spaces are struc-tured along similar lines, as are representations of external space,and there are natural links between the two. This provides a novelway to link perceptual and physical spaces.

Urban versus Suburban Location, Environmentand BehaviorThe space of a city is composed of an overlap of different landcovers, land uses, and environmental characteristics. The urbanarea is characterized by a great density of buildings and, inversely,a scarcity of green spaces. The characteristics of this environmentare one of the determinants in the choice of residential location.With simplification, urban inhabitants appear to have greaterpreference for the urban characteristics of the environment thando the inhabitants of suburban regions, who appear to prefergreenness, nature, and a greater house or lot size. Environmentalvariables are certainly important to characterize the different homespaces in a city. We can call this the “environmental description”or quality of the human space.

These spaces are used differently by people as a function oftheir activities, their environmental sensitivity, and their social char-acteristics. Pacione (1990) introduced the concept of urban live-ability, a relative term whose precise meaning depends on the place,time, and purpose of the assessment and on the value system of theassessor. Pacione emphasized the possible delimitation of sub-areasin the city based on the relationship between environment andsocial behavior. Remy and Voye (1993) have suggested a differen-tiation of spatial perception according to the spatial behavior ofgroups in situ. Activity patterns help determine the extent of per-ceived neighborhoods (Walmsley and Lewis 1993) and contributeto the “environmental structure” of the human space. It is likely,for example, that the environmental structure of space for personswho exhibit urban behavior is smaller and closer to their homesthan for persons who exhibit suburban behavior.

However, there are two explanatory factors for such patternsof activities: the household (or family context) and the spatial

context. Accessibility issues will impact the patterns of everydayactivities less in urban areas than in suburban areas, because ofthe density of places where activities may occur (Hanson 1982).Due to the variation among environmental characteristics, wecan assume that, in general, the spatial context in urban or sub-urban areas will influence the structure of the individual life spacesof the inhabitants. More precisely, location affects the modes andfrequencies of everyday mobility, the way in which knowledgeabout the environment is acquired, and the presence of opportu-nities and constraints in the local environment.

ObjectivesFrom a review of the literature, that it should be possible to de-velop hierarchical spatial units that are representative of spatialbehavior of people. Indeed, the shape of these units and the pres-ence of additional environmental constraints constitute assess-ments of the environmental sensitivities of people. Although thesurvey shows that perception has been widely recognized to playa role, no single method for handling perception from a method-ological perspective has emerged.

The objectives of this article are to propose a clear method-ology for including perception in the analysis of environmentalsensitivity of households and to explore the implementation ofthis methodology in a geographic information system (GIS) struc-ture. In particular, we introduce the concept of perceptual re-gions, a new hierarchical set of spatial units that join location toactivities, the two key concepts of life spaces. The structural andenvironmental differences of the perceptual regions are exploredin relation to the sociological characteristics of urban and subur-ban behaviors.

In this article, the concept of perceptual regions is intro-duced. These are comprised of three hierarchical spaces whereperception plays different roles. Following this, the methodologyand hypotheses of implementation are presented. An analysis ofenvironmental quality is then presented for two districts withinthe city of Québec (Canada) with different urban characteristics.The data are derived from remotely sensed images, topographi-cal maps, and an origin/destination survey that provides infor-mation about the activity spaces. Discussion concerningenvironmental sensitivity in relation to urban and suburban resi-dential choices and behaviors follows these results.

The Concept of Perceptual RegionsThe concept of a perceptual region addresses the integration ofthe two key concepts in each of these analyses: location and ac-tivities (Garling and Golledge 1993). Associated with a location(or place) are a set of qualities and a set of activities carried out byagents within the location (Figure 1). The combination of theseperceived qualities (to the extent that they are viewed in a posi-tive light) and the activities undertaken generate a sense of be-longing for the agents with respect to this location. These activitiesare everyday experiences carried out in and around one or moreplaces. Furthermore, the everyday activities that occur within theseregions enter into the agents’ memory and are reinforced or for-


gotten as a function of how often the activities are repeated. Hence,frequency is a major determinant for which parts of an environ-mental space are perceived by individual agents. Different scalesof perception correspond to different subsets of these activities.

The Perceptual RegionsThe perceptual region is comprised of three elements in interac-tion: 1) a sense of belonging to a space, associated with a hierar-chical structure; 2) a set of environmental qualities; and 3) acollection of activities (the perceptual regions are characterizedby nodes and throughways where activities converge). Further-more, the perceptual regions embrace at least three types of spaceor sub-region, which are denoted here as the vista space, the localdisplacement space, and the enlarged displacement space.

Vista SpaceVista space is the first space of the hierarchical structure and cor-responds to the spatial node of the activities. It is based on thenotion defined by Montello (1993): “as large or larger than thebody but can be visually apprehended from a single place with-out appreciable locomotion. It is the space of single rooms, townsquares, small valleys, and horizons.”. We adapted this definitionto the context of the evaluation of environmental quality: it is aspatial region with perceptually similar characteristics appre-hended from a single place, but not determined by vision alone,and which corresponds to a sense of belonging resulting fromactivities carried out in that region. The vista space can be resi-dential (and, hence, related to the concept of a home), a workingplace (for adults) or school (for children), or places for other ac-tivities. It is the space where the majority of variables of environ-mental quality are assessed (Reginster and Edwards 2001).

Local Displacement-Reinforcement SpaceLocal displacement space is the space surrounding the vista spaceand which comprises the place of frequent visits. To define it, weuse the notion of reinforcement.

In related work, Edwards (1997, 2000) developed softwarearchitecture for simulating spatial memory. The model used bythis simulation is based on the principle that knowledge about aspatial environment is obtained by traversing it many times. Thereinforcement of matched events across different trajectories isused to build a number of representations of the space traversed.These representations include lists of reinforced events and eventaggregates (chains of events, objects, group, and sites), as well asnetworks of sites and objects constructed from the trajectories.Distance estimates are likewise managed within the simulationvia a set of independent algorithms for measuring distance thatare combined into a weighted average. These algorithms includemeasures of subjective time, visual (angular) measures of distance,and the use of a characteristic distance associated with each ag-gregated event. The characteristic distance represents the typicalseparation of objects of a similar type. Using these distance esti-mates combined with a network representation of sites, it is pos-sible to construct a map representation that assigns a characteristicspatial region to each object. Associated with each object is areinforcement or importance rating in addition to this character-istic size. This representation is congruent with a similar deviceused by Kettani and Moulin (1999) to develop a reasoning en-gine for planning a route through an arbitrary cityscape.

This representation can be used to define the displacementspaces. The reinforcement rating is based on the frequency withwhich the location is incorporated into a household’s daily jour-neys, and the spatial region is a circle or ellipse that characterizesthe spatial separation of the typical object at the location con-cerned with respect to other objects of the same type. The unionof the set of overlapping regions with reinforcement above some

Figure 1: The concept of perceptual regions and the relationships between thematic elements addressed.


Figure 3: The residential vista space for Sillery, Québec (part of aninfrared aerial photo, originally obtained at 1:15,000 scale).

Figure 2: Hierarchical structure of the perceptual region of a district.Shown are vista spaces (V.S.) for residences, work, and other locations,local displacement spaces, and enlarged displacement spaces.

Figure 4: The local displacement space (part of the road network) forSillery, Québec. Roads are shown in red, autoroutes in dark red,autoroute exits in orange, commercial zones in dark blue, schools inblue, and green spaces in yellow.

Figure 5: The enlarged displacement space (part of an origin-destination survey) for Sillery, Québec. All linear elements show actualdisplacements along the road system. The thicker and darker the line,the more frequent the displacements along that segment of road andthe more heavily reinforced that element of the space.


sponds to the sense of belonging to the immediate environmentaround an activity center such as the “home.”

An approximation of the local displacement space could bestructured by the use of locomotion inside reasonable temporallimits; however, this has several undesirable characteristics. It leadsto a sharp cut off between what is considered to be located withinthe local displacement space and what is outside this space. It isalso subject to the need to define the means of locomotion, andhence might, in principle, lead to several local displacement spaces.However, in the absence of appropriate data in the form of ac-tivities in the immediate neighborhood of the residence or otheractivity locus, such an approximation may be readily developedusing available cartographic data.

Enlarged Displacement-Reinforcement SpaceThe enlarged displacement-reinforcement space consists of theregion that embraces both the activity islands beyond the localreinforcement region and the local reinforcement region itself.The enlarged displacement-reinforcement space is hence more anetwork-like space than an areal space (i.e., it is full of holes!).

Hierarchically Structured Perceptual RegionsThe perceptual region, as we define it, includes a hierarchicalstructure, with a succession of spaces embedded in others, to whichpeople may have a sense of belonging or attachment. Hence, vistaspaces are embedded in local displacement spaces and the latterare embedded in enlarged displacement spaces (Figure 2). Largerspaces may also be envisioned (e.g., geographical space as de-fined by Montello 1993), but are not needed for the purposes of

Figure 6: The residential vista space for Charlesbourg, Québec (part ofan infrared aerial photo originally obtained at 1:15000 scale).

Figure 7: The local displacement space (part of the road network) forCharlesbourg, Québec. Roads are shown in red, autoroutes in darkred, autoroute exits in orange, commercial zones in dark blue, schoolsin blue, and green spaces in yellow.

Figure 8: The enlarged displacement space (part of an origin-destination survey) for Charlesbourg, Québec. All linear elementsshow actual displacements along the road system. The thicker anddarker the line, the more frequent the displacements along thatsegment of road and the more reinforced that element of the space.

level can then be used as a map of the displacement space. In thefirst definition, the local displacement-reinforcement space willcorrespond to the region surrounding the home; it maps the lim-its of what is well integrated into the household’s activities. Eachspatial location (e.g., house, street, park, or school) can be as-signed a reinforcement rating and a characteristic spatial region.The spread of the reinforced zones around the home may be usedto calibrate the scale of the local displacement region used toapproximate the reinforcement zone. This space, like the vistaspace, can be delimited by barriers in several directions. It corre-


this study. For the most part, the activities that take place in vistaspace are household or workplace activities, while activities thattake place in displacement spaces are locomotive and broader inscope. Vista spaces will be heavily reinforced, local displacementspaces somewhat less so, and enlarged displacement spaces evenless reinforced. There are clearly even larger regions that incorpo-rate occasional displacements and perceptual regions that are ac-quired via information sharing with the perceptual regions ofother agents. Another characteristic of the displacement spaces,highlighted by their origin in memory reinforcement, is that theyare heterogeneous regions. They contain large holes where noinformation may be available (frequentation is nil).

Implementation and Tests on Urbanand Suburban RegionsThe concept of perceptual regions is implemented in a GIS fortwo districts within the city of Québec (Canada). These districtsare homogeneous residential districts largely characterized bysingle-family dwellings. The definitions of the three spaces thatcompose the perceptual region are aggregated for the set of theinhabitants of each of the two districts. The areas chosen for theanalysis within Charlesbourg and Sillery are homogeneous bothin population and from a morphological point of view.

The district of Sillery, located in the western portion of thecity, is closer to the centers of work (e.g., university, administra-tions, and shops) than the district of Charlesbourg. The area ofanalysis comprises 1.19 sq km and 1776 inhabitants. Figure 3shows the an aerial view of the region used in the study to definethe vista space, Figure 4 shows the cartographic data used to de-fine the local displacement space, and Figure 5 shows the pathtraces of displacements derived from the origin-destination sur-vey and, hence, constitutes a visualization of the enlarged dis-placement space of the neighborhood.

The district of Charlesbourg, located in the northern part ofthe city, is typically a suburban district. The studied part of thedistrict comprises 4902 inhabitants who for the most part live inbungalows or cottages. The study area covers 2.23 sq km (seeFigures 6, 7, and 8, which are comparable to Figures 3, 4, and 5for the Charlesbourg district).

DataThe implementation requires two kinds of data: 1) spatial cover-age including environmental variables, amenities, and constraints;and 2) information about the displacements and locations of ev-eryday activities of family members of the households of the dis-trict.

The first type of information is partially derived from re-motely sensed images and aerial photos. Greenness and sourcesof nuisances are extracted from these data. Information frommaps, such as the location of amenities, is acquired to completethe spatial coverage. The vista spaces for both districts were ex-tracted by determining the presence of visual barriers for a sub-sample of households in each district.

Information about the displacements comes from an origin-destination survey of individual daily mobility organized bySTCUQ, the public transportation company of the city of Québec(Theriault et al. 1996). This survey was done by telephone in1991, for a population sample corresponding to 8.2% of the to-tal population of the districts. The survey is used to assess theenlarged displacement space of inhabitants. Because of the typeof survey and its goal, it was not possible to assess the local dis-placement-reinforcement space in the same way. Hence, we usethe approximation based on the temporal limits of walking, asoutlined earlier (Figures 4 and 7).

The data used and extracted from the survey concern:� 128 households of the Charlesbourg district, which corre-

spond to 402 inhabitants (8.2% of the total population)and 984 displacements (2.4 displacements/inhabitant).

� 51 households of the Sillery district, which correspond to142 inhabitants (8.2% of the total population ) and 296displacements (2.1 displacements/inhabitant).

The Perceptual Spatial Region of the DistrictsThe perceptual spatial region of a district comprises three spaces:� the residential vista space aggregated for the inhabitants of

the district;� the local displacement space aggregated for the inhabitants

of the district and which could include non-residential vistaspaces such as those for shopping centers; and

� the aggregated enlarged displacement space that includes andlinks the set of vista spaces of the inhabitants of the district.

Hypothesis and Implementation of the Percep-tual Region of Each DistrictThe identification and implementation of the perceptual regionin an urban or suburban area require some additional hypoth-eses, particularly concerning the manner in which people definetheir sense of belonging in terms of these hierarchical spaces.

The most frequented space is usually the residential vistaspace. In the suburban region, the aggregated residential vistaspace is the set of homes and the green space surrounding it thatis viewable from the homes even if not within the property of thehousehold. This space can be limited by barriers such as mainroads, autoroutes, reliefs, or walls. In suburban areas, this space isgenerally characterized by the presence of greenness. In urbanareas, because of the proximity of buildings, the barriers are gen-erally closer to the houses and the residential vista spaces are lessgreen than in suburban areas. The identification of this spacewith fuzzy limits is possible on infrared aerial photos at the1:15,000 scale. Hence, it is assumed that the perception of theenvironment is strongly determined by the sense of attachmentto this vista space, and, in particular, to its greenness.

The local displacement space is assessed on the road andpath network. The second definition of this space is used forimplementation, characterized by a temporal limit of locomo-tion by foot. The limits used are 5 or 7 minutes (500 m) of a


normal walk by foot; this region is seen in Figure 4 for Silleryand Figure 7 for Charlesbourg). It is not expected that the resultswill depend heavily on small differences in the choice of this limit.To implement this second space, a hypothesis on the relation-ship between sense of belonging and the region characterized bythe duration of a short walk is posed.

The enlarged displacement space is assessed from the ori-gin-destination survey. It corresponds to the space frequented bypeople in a single day. Figure 5 shows the origin-destination dataset graphically for Sillery, while Figure 8 shows the correspond-ing diagram for Charlesbourg. The lines shown correspond todisplacements revealed by the survey. Thicker and darker linesare more frequent (i.e., reinforced) displacements. The identifi-cation of this third space exploits the hypothesis of a relationshipbetween the sense of belonging to a space and those parts of thespace actually visited.

The proposed implementation of the theoretical conceptsdeveloped earlier clearly represents a certain simplification. A morecomplete data set might use a digital elevation model with a de-tailed micro-relief to determine the vista space, as well as a surveythat specifies displacements by foot and vehicle. However, thedata set used for this study, along with these simplifying assump-tions, appeared to be effective as a test bed for the methodologydeveloped and proposed here, as can be judged from the resultsreported below.

From a first visual analysis of the component of the percep-tual region of Charlesbourg (Figure 6), it is possible to underlinethe importance of the greenness inside the residential district.The residential vista space and the local displacement space arecharacterized by the presence of an important barrier and sourceof nuisances: an autoroute. The enlarged displacement space de-rived from the origin-destination survey presents a great varietyof destination spaces with a general main direction to the south,where many of the work places of the city are localized.

The part of the Sillery district studied (Figure 3) is also char-acterized by a great deal of green. The road network is limited onthe south by the presence of a cliff (Figure 5). The enlarged dis-placement space is more oriented than that of Charlesbourg. It ispossible to identify one main direction of displacement to theeast, toward the center of the city.

Development of the Tests Inside the PerceptualRegionsThe validation of these perceptual regions is carried out by assess-ing environmental variables inside each space. The goal of the analy-sis is to test the environmental sensitivities of two different districts:a suburban area and an area closer to the center of work.

The environmental variables are developed from remotelysensed data and via geomatics tools. These are adapted accordingto the different structural elements of the vista space, the localdisplacement space, and the enlarged displacement space.

The suggested environmental variables are:� the size of the spaces in relation to the number of dwellings,

which is a measure of density of the district;� the greenness of the space which includes the gardens and

public green spaces, an indicator of ecological quality;� the presence of nuisances determined via the area affected

by the negative influence of factories or highways (othersources of nuisances can include garbage dumps, quarries,an airport, etc.);

� the presence of barriers determined via the number of barri-ers for pedestrian or car locomotion according to the defini-tion of the spaces; it is important to note that the definitionsof barriers are different in vista space (view-based), in localdisplacement space (pedestrian definition), and in enlargeddisplacement space (car locomotion definition);

� accessibility is measured as a percentage of the area coveredby public services (e.g., schools and shops) in the local dis-placement spaces or as the average distance inside the en-larged displacement space;

� the percentage of short distances (e.g., below a threshold)compared to all distances and to destinations inside the en-larged displacement space acts as an indicator of the life ofthe district;

� the percentage of long distances (e.g., above a threshold)compared to all distances and to destinations that are out-side the zone of attraction;

� the length of the largest dimension for the enlarged displace-ment space per inhabitant, which is an indicator of lengthof the displacement of each inhabitant; and

� the length of the largest dimension for the enlarged displace-ment space not affected by other districts can be used tomeasure the displacements that are unique to the inhabit-ants of the given district.

Using some variables, it is possible to identify or establishwhat could be considered the best value by most people in termsof quality. A large value of greenness or a low value of nuisancescould be identified as more desirable in accordance with ecologi-cal principles. Moreover, the presence of services could be con-sidered as more desirable in accordance with accessibility priorities.It is important to note that some variables could have a doubleimpact: positive in term of accessibility, but negative in term ofnuisances. This is the case of the accessibility to an entrance to anautoroute, which could have noise effects. Some other variablesare more dependent on the sensitivity of inhabitants. This is thecase, for example, with the size of the vista space, which is typi-cally representative of an attachment to urban or suburban sensi-tivities. A large vista space could be important for the preservationof private life, but could be considered as a social gap for others.From this set of variables, it is possible to develop a general indi-cator of environmental quality of the district or an interactiveone in accordance with the priorities of individual persons.


Figure 9: Oriented diagram of displacements for Charlesbourg and Sillery.

Table 1: Comparison of the Two Districts with Environmental Indicators Assessed in the Perceptual Regions

Charlesbourg Sillery(suburban district) (closer to jobs)

Vista space size, normalized to a single household 0143 ha 0191 haGreenness of the vista space (maximum = 1) 0.31 0.44Presence of nuisances in the vista space (maximum = 1) 0.19 0.01Presence of barriers in the vista space 2 0Local displacement space size, normalized to one household 0,227 ha 0,485 haGreenness of the local displacement space (maximum = 1) 0.28 0.43Presence of schools inside the local displacement space 7 2Percentage of the area of shops inside the local displacement space 1.8% 0.4%Percentage of the area of parks inside the local displacement space 4.1% 2.3%Presence of an entrance to an autoroute inside the local displacement space 2 0Presence of barriers in the local displacement space 1 0Average distance to a work destination (origin-destination survey) 7.47 km 5.94 kmAverage distance to a school destination (origin-destination survey) 3.96 km 2.47 kmAverage distance to a shop destination (origin-destination survey) 3.56 km 3.63 kmPercent of short-distance time (<5 min) to destinations (indicator of district life) 34.2% 53.9%Percent of long-distance time (>25 min) to destinations 2.5% 5.3%Length of the through-space/number of inhabitants questioned 16,161 m/inhab. 10,4


Results of Environmental SensitivityTests in Perceptual Regions

Comparison of the Two Districts with Environ-mental Variables

Table 1 compares the two districts in accordance with envi-ronmental variables assessed and adapted inside the three hierar-chical spaces. A comparison of the two districts permits a betterunderstanding of the environmental sensitivities of the inhabit-ants of the two districts.

� The district of Charlesbourg is characterized by a lower valueof greenness than Sillery, although the values indicate a com-fortable level of greenness for both districts.

� Charlesbourg is clearly affected by more nuisances and bar-riers than is Sillery.

� The accessibility variables of both districts are somewhatbalanced. Charlesbourg includes much more service infra-structure than does Sillery. The accessibility assessed in theenlarged displacement space, however, is more favorable forSillery, with lower average distances to work and school des-tinations. This last example underlines the importance ofthe spatial units used for assessment.

� The size of the local displacement space, normalized for thenumber of households, is higher in Sillery than inCharlesbourg because of the presence of barriers in the lat-ter.

� The indicators of quality of life in the district extracted fromthe origin-destination survey present a better value for Sillery(53.9% of short-distance displacements versus 34.2%). How-ever, Sillery is also characterized by a higher percentage oflong-distance displacements (5.3% versus 2.5%).

It is important to note the relationship between environ-mental variables and home prices, as well as their influences onthe residential choices of individuals. The environmental vari-ables in the Sillery district are in general more advantageous interms of quality; however, the average cost of housing is muchhigher than for Charlesbourg. The hedonic price is an econo-metric technique that allows for assessment of the implicit pricesthat people may be expected to pay for a given environmentalfactor or amenity. It allows the introduction of variables of com-fort or house size as well as environmental variables, treated asindependent variables in multiple regressions. The method wasdeveloped by Rosen (1974). Numerous empirical tests of thesemethods exist. Improvements have been made and carried outfor detailed environmental variables by Reginster and Goffette(2001) and Reginster and Edwards (2001).

Descriptive Diagrams to Compare the DistrictsThe environmental functioning of a district or the environmen-tal sensitivity of its inhabitants could also be assessed by the use

of diagrams. The one suggested here concerns the enlarged dis-placement space and is derived from the origin-destination sur-vey. The diagram in Figure 9 reflects the orientation of thedisplacements of the inhabitants of a district. It allows a descrip-tion of the major directions of displacements, of the presence ofbarriers in the enlarged displacement space, the directions andmeasurements of distance to the centers of work, and an indica-tor of the life of the district derived from the percentage of dis-placements within the district.

In Figure 9, “mt” refers to the average time of displacement

(in seconds/100), and the direction sign indicates the directionof the city center (center of work). A comparison of the two dia-grams representing the enlarged displacement spaces allows theidentification of a main barrier in the southeast direction forSillery, which corresponds to a cliff and the St, Lawrence River.No such barrier is derived from the graphic for Charlesbourg,even if there are fewer displacements in the northwest direction.Charlesbourg appears to have a weaker district life than Sillery, asindicated by the percentage of displacements inside the district(44% versus 50%). This was mentioned in the previous analysisbased on a different variable (the number of short-distance dis-placements).

Another diagram representative of the life of the district is afrequency diagram: the frequencies of the displacements as a func-tion of their duration. The natural breaks in such a representa-tion allow for the identification of thresholds for environmentalvariables as was done in Table 1 (the percentage of short-distancedisplacements to destinations within the enlarged displacementspace for a duration of less than 5 minutes, and the percentage oflong-distance displacements to destinations for a duration of morethan 25 minutes).

Discussion and ApplicationsThese methodologies and implementations could be enlarged toinclude all the districts of the agglomeration. They also can beadapted for other types of variables or by changing the level ofanalysis. One way to quantitatively present the environmentaldistinctions between the districts is to perform an analysis of vari-ance. This method allows measurement of the separability of thedistrict and, in this case, of the environmental quality. The pro-cedure is hard to justify for the analysis of only two districts, butcould be reasonably applied to a more thorough study of theentire city.

The values presented in Table 1 reflect only some aspects ofthe complexity of these two districts. Additional data could beobtained by obtaining a variety of other factors, such as, relativevalues of road safety or management. Such variables could in-clude the percentage of the enlarged displacement spaces that areassociated with highways or the presence of dangerous roads orcrossroads in the vista spaces. These variables could be used toaddress applications relative to habits of transportation, trafficcirculation schema, or public transportation policies.

The data derived from the origin-destination survey con-cern the displacements of persons during a single day. In the case


of another survey, the implementation of local and enlarged dis-placement spaces could be adapted from, for example, a weeklysurvey. This would provide more information concerning rein-forcement of activity sites and should support a finer segmenta-tion of household groups. This would also obviate the need toresort to approximations for the local displacement space, suchas the 5- or 7-minute limit used in this study.

Although developed in this article at the spatial level of thedistrict, the perceptual region can also be implemented at thelevel of each house. This individual level requires other hypoth-eses and techniques of analysis, and has other applications(Reginster and Edwards 2001).

Applications for these perceptual regions used as spatial unitsfor analysis include characterizations of residential choice (seeReginster and Edwards (2001) for more details of the latter),closely tied to the analysis presented in this article. However, stud-ies in travel and activities planning and retail choice may profitfrom the hierarchical nesting of the regions proposed. For ex-ample, the Huff model (1962) could be adapted to include per-ceptual reinforcement and applied using perceptual regions fordifferent client groups. Other applications under study includeintegrated forest management, the development of landscapepattern metrics that incorporate perceptual factors, and the de-sign of natural and urban park spaces.

In a general way, these spatial units could become the basisof new indicators of environmental factors such as quality. Thetypology of the structure of the perceptual regions could lead toa typology of the districts of a city as determined by the spatialbehaviors of their inhabitants.

ConclusionsIn this article, we have developed new spatial units that incorpo-rate the cognitive and perceptual characteristics of human at-tachment to places. These spatial units could become the basis ofnew indicators of environmental factors for a variety of applica-tions concerned with environmental sensitivity. The definitionand identification of the perceptual regions are based on hypoth-eses concerning the importance of life spaces for a set of house-holds. These hypotheses allow the integration of different typesof human spaces: areal for vista spaces and network-like for thedisplacement spaces. These are a reflection of the different activi-ties and different kinds of attention people give to their environ-ment. Moreover, the hierarchical structure allows the adaptationof different measures to each of the spaces within the perceptualregions.

Also described in this article is a method for extracting theseperceptual regions from appropriate data, including remotelysensed imagery, digitized cartographic data, and origin-destina-

tion surveys; the manner in which these regions may be used isshown as structural units for aggregating spatial data appropri-ately for studies of human spatial behavior. Perceptual regionsappear to offer a useful means of aggregating spatially distributeddata in support of statistical analyses for a variety of applications(see Reginster and Edwards (2001) for a description of other typesof analysis that may be carried out using perceptual regions).

Instead of using spatial units based on political divisions(district boundaries or census tracts), the proposed approach al-lows one to transform measures to correspond to the environ-ments perceived by inhabitants of the territory (assumingdisaggregated data are available to support this method). We havedeveloped such an analysis by aggregating data across a neigh-borhood; however, the same kind of approach could be carriedout based on other aggregation criteria. For example, childrencan be expected to have a different perception of environmentalquality than their adult counterparts. The difference is partly dueto the fact that the criteria they consider important are different,but also because the territory of which they are conscious is quitedifferent than that perceived by adults.

This study is an attempt to formalize a certain understand-ing of the structure of residential space, with a comparison ofdifferent sizes of space. The approach offers the potential for de-veloping some interesting applications in urban planning: themeans to be more sensitive to the wishes of individual house-holds in decisions concerning urban spaces, and a tool to helppeople evaluate different residential locations. More globally, thisstudy constitutes an experiment in the comprehension of the re-lationship between behavior, residential structure, and environ-ment.

Finally, it is worth pointing out that the analysis undertakenin this study can be largely automated. The extraction of vistaspaces requires the combined use of remotely sensed imagery giv-ing access to greenness measures and of view shed analysis basedon digital elevation models. In this study, no significant viewshed analysis was performed because the regions studied are flat.The origin-destination surveys can likewise be analyzed fairlyautomatically. As a result, it is possible to perform these kinds ofanalysis on large data sets such as major urban agglomerationsand still obtain information at the levels of detail of individualhouseholds, provided that the supporting data exists (maps andorigin-destination surveys).

URISA Journal � Butler, Dueker 15

About the Authors

Isabelle Reginster received a PhD from the University of Louvain(Belgium). She spent one year at the Centre de Rechercheen Geomatique (Universite Laval-Canada) where she hasachieved this research. Her current research interests are inthe urban environment, sclae and spatial units and land coverchanges. She may be contacted at the Department of Geog-raphy, Universite de Louvain, Place Pasteur, 3, 1348,Louvain-la Neuve, Belgium. Tel: 32-10-472869. Fax: 32-10-472877. [email protected].

Geoffrey Edwards holds a B.Sc. in Astronomy (1979) from theUniversity of Victoria, and an M.Sc. (1982) and a Ph.D.(1987) in Astrophysics from Laval University. He has beenworking in the Department of Geomatics Sciences at LavalUniversity since 1987. He is a founding member of the Cen-tre for Research in Geomatics, currently holds the rank ofFull Professor and is Associate Program Leader for aCanada-wide geomatics research network called GEOIDE.He has recently been awarded one of Canada’s prestigiousResearch Chairs. His research interests lie at the interfacebetween geomatics and cognitive science and are targetedtowards developing better tools and methods in applicationsas widely ranging as vision, navigation, planning and de-sign. The author may be contacted at The GEOIDE Net-work, Pavillon Casault, Université Laval, Sainte-Foy, QuébecG1K 7P4, Tel: (418) 656-2196. Fax: (418) 656-2611. email:[email protected].

Acknowledgments

We would like to thank Dr. Marius Theriault for kindly pro-viding us with access to the origin-destination survey developedin conjunction with the CTCUQ. Funding for this project wasprovided in part by the Industrial Chair in Geomatics applied toforestry and hence by the Québec Association of Wood Sawmillsand Manufacturers (AMBSQ), the Ministry of Natural Resourcesof Québec (MRNQ), the Natural Sciences and Engineering Re-search Council (NSERC), and Rexfor. The final stages of thiswork were funded by the GEOIDE Network.

References

Adams, J.S., 1984, The Meaning of Housing in America. Annalsof the Association of American Geographers, 74, 515-526.

Aitken, S. K., and R. Prosser, 1990, Resident’s Spatial Knowl-edge of Neighbourhood and Form. Geographical Analysis,22, 301-325.

Crevoisier, O., 1996, Proximity and Territory versus Space inRegional Science. Environment and Planning A, 28, 1683-1697.

Davis, E., 1981, Organizing Spatial Knowledge. Publication #193,Computer Science Department, Yale University, New Ha-ven, CT.

Edwards, G., 1997, Geocognostics – A New Framework for Spa-tial Information Theory. In A.U. Frank and S. Hirtle (Eds.)Lecture Notes in Computer Science, COSIT’97, (New York:Springer-Verlag), 455-471.

Edwards, G., 2000, A Memory Architecture for Simulating Spa-tial Cognition. In preparation.

Fotheringham, A.S., and A. Curtis, 1992, Encoding Spatial In-formation: The Evidence of Hierarchical Processing. In A.U.Frank, I. Campari, and U. Formentini (Eds.) Lectures Notesin Computer Science, COSIT’92, (New York: Springer-Verlag), 269-287.

Gale, N., R.G. Golledge, W.C. Halperin, and H. Couclelis, 1990,Exploring Spatial Familiarity. The Professional Geographer, 42,299-313.

Garling, T., and R.G. Golledge, 1993, Understanding Behaviorand Environment: A Joint Challenge to Psychology andGeography. In T. Garling and R.G. Golledge (Eds.) Behav-ior and Environment, Psychological and Geographical Ap-proaches, (Netherlands: Elsevier Science Publishers), 1-15.

Golledge, R.G., and R.J. Stimson, 1997, Spatial Behavior – AGeographic Perspective, (New York: Guilford Press), 620 pp.

Hirtle, S.C., 1995, Representational Structures for CognitiveSpace: Trees, Ordered Trees and Semi-Lattices. In A.U. Frankand W. Kuhn (Eds.), Lecture Notes in Computer Science,COSIT 95, (New York: Springer-Verlag), 327-340.

Hirtle, S.C., and J. Jonides, 1985, Evidence of Hierarchies inCognitive Maps. Memory and Cognition, 13(3), 208-217.

Hanson, S., 1982, The Determinants of Daily Travel-ActivityPatterns: Relative Location and Socio-Demographic Factors.Urban Geography, 3, 179-202.

Huff, D.L., 1962, Determination of Intra-urban Retail Trade Ar-eas, Real Estate Research Program, University of California,Los Angeles.

Kettani, D., and B. Moulin, 1999, A Spatial Model Based on theNotions of Spatial Conceptual Map and of Object’s Influ-ence Areas, Lecture Notes in Computer Science, COSIT 99,(New York: Springer-Verlag), 401-416.

Lyon, L., 1987, The Community in Urban Society, (Chicago:Dorsey Press).

McNamara, P., 1986, Mental Representations of Spatial Rela-tions. Cognitive Psychology 18, 87-121.

Montello, D.R., 1993, Scale and Multiple Psychologies of Space.In A.U. Frank, I. Campari, and U. Formentini (Eds.) Spa-tial Information Theory, COSIT 93. Lectures Notes in Com-puter Science, (New York: Springer-Verlag), 312-321.

Norberg, and C. Schultz, 1971, Existence, Space and Architecture(New York: Praeger), 120 pp.

Norcliffe, G., and L. Xiaofeng, 1993, Home, Work, and Local-ity/The Case of Thorold, Ontario. Urban Geography, 14, 251-276.


Pacione, M., 1990, Urban Liveability: A Review. Urban Geogra-phy, 11, 1-30.

Reginster, I., and F. Goffette, 2001, Environmental Quality andResidential Choices. Submitted.

Reginster, I., and G. Edwards, 2001, Development of SpatialRepresentations of Environmental Quality Based on Percep-tual Regions. Poster Session of COSIT ‘99.

Reginster, I., and G. Edwards, 2000, Assessment of the Environ-mental Quality inside Perceptual Regions using Economet-ric Tests. Submitted.

Rosen, S., 1974, Hedonic Prices and Implicit Markets: ProductDifferentiation in Pure Competition. Journal of PoliticalEconomy, 82, 34-55.

Shanton, B., 1986, The Incidence of Home Grounds and Expe-riential Networks. Environment and Behavior, 18, 299-329.

Taylor, R.B., 1988, Human Territorial Functioning, (Cambridge:Cambridge University Press), 351 pp.

Theriault, M., C. Thomas, and M. Beaudoin, 1996, Fichiers desDéplacements de L’enquête O-D 1991 de la STCUQ.

Tuan, Y.-F., 1996, Space and Place: A Humanistic Perspective. InG. Agnew, D.N. Livingstone, and A. Rogers (Eds.) HumanGeography, An Essential Anthology, (London: Blackwell Pub-lishers), 444-457.

Waddell P.A. , and V. Shukla, 1993, Manufacturing Location ina Polycentric Urban Area: A Study in the Composition andAttractiveness of Employment Sub-centers, GeographicalAnalysis, 14:3, 277-296.

Walmsley, D.J., and G.J. Lewis, 1993, People and Environment.Behavioral Approaches in Human Geography,. (New York:Longman Scientific and Technical), 290 pp.

The Urban and Regional Information Systems Association (URISA) is the premier

professional association for those involved in improving our urban and regional

environments through the effective use of information technology. Professionals in

planning, economic development, information systems, emergency services, natural

resources, public works, transportation, and other departments within state and local

government have depended on URISA for professional development and educational

needs since 1963. Through its international, national and local chapter operations,

URISA serves nearly 8,000 professionals.

What is URISA?


IntroductionThe development and use of geographic information systems(GISs) have been areas of activity in transportation agencies formany years. Yet, significant progress toward realizing the prom-ise of GIS as an integrator of data and operations in these agen-cies has been rare. Legitimate differences in requirementsfrequently lead to application-specific definitions and represen-tations of transportation features and their geometry. For ex-ample, vehicle navigation systems, overweight truck routing,facility location, address geocoding, and emergency manage-ment applications all have unique requirements. The challengeis to establish a means of data exchange among these disparaterepresentations, one that leads to overall improvements in ac-curacy, consistency, and completeness as the users of each sys-tem perceive these qualities.

Deploying GIS in transportation (GIS-T) at an enterpriselevel presents an opportunity to eliminate the traditional appli-cation-specific development pattern of information systems byproviding a common data structure centered on transportationfeatures. The solution is to embrace the diversity of applicationsand data requirements within a unifying enterprise data modelfor GIS-T that allows each application group to meet the estab-lished needs while enabling the enterprise to integrate and sharedata. The primary objective of this model is to allow frequenttransaction-based data exchanges and updates, the type that aninteractive organization is likely to need. As transportation agen-cies move toward a more integrated manner of doing business,such as involving design units earlier in the project planning cycle,the need for data to cross former institutional or jurisdictionalbarriers will become greater.

OverviewThis article will show how a GIS-T database can be constructedto facilitate data exchange. The ultimate goal is to construct asingle database that will be used by all applications without theneed for formal data exchange mechanisms. The design focuseson a transportation feature table that implements the conceptsexpressed by the authors in earlier articles (Butler and Dueker

Implementing the Enterprise GIS in TransportationDatabase Design

J. Allison (Al) Butler and Kenneth J. Dueker

Abstract: Earlier articles by the authors described the primary database design approaches that have been and are being used ingeographic information system applications for transportation. These articles proposed a theoretical model for an enterprisegeographic information system environment for transportation agencies. This follow-up article provides detailed guidance onbuilding a multimodal transportation feature database using relational database concepts.

2000, 1998, Dueker and Butler 1999, 1998). These articlespresent theoretical and implementation aspects of an EnterpriseGIS-T Data Model that includes four primary components:

1. a facility inventory comprising jurisdiction, transportationfeatures, event points, linear events, point events, and inter-sections;

2. a network that includes nodes, links, traversals, and traversalsegments;

3. a measurement datum consisting of anchor points, anchorsections, reference objects, and geographic points; and

4. cartography, which includes (in the simplified version dis-cussed here) base map strings, linear event strings, line seg-ments, point symbols, and cartographic points.

The list of components is essentially an “all of the above”response to the existing structures of legacy transportation infor-mation systems and current GIS product lines (Butler and Dueker2000). At the same time, the unbundled approach of the systemallows system designers to pick the elements that are needed for aparticular application (Dueker and Butler 1998). For example,Want to draw a map of an existing highway inventory? It is nec-essary to:

1. Choose a GIS platform that supports dynamic segmenta-tion.

2. Establish a one-to-one relationship between each transpor-tation feature map object and its corresponding feature inthe inventory. This is accomplished by constructing the ob-ject and then attaching the transportation feature identifierin the inventory database.

3. Assign measures for beginning and ending points for linearobjects representing linear transportation features.

4. Use the dynamic segmentation function to create a carto-graphic object for each linear event or to position pointevents.


Another example would be to route an oversize/overweighttruck. Assuming that the facility inventory has all the data neededto select route segments, the steps listed above will need to bedone as well as the following:

1. Construct a node table of all points where a route segmentcan begin, end, or change from one transportation featureto another. Each node must have a unique identifier.

2. Create a link table containing all of the valid links. Eachlink must have a unique identifier and the related recordmust include the terminating node identifiers. An ordered-pair approach allows the ability to distinguish the directionof travel.

3. Extract a set of linear and point event table records contain-ing relevant attributes, such as bridge width, height, andload capacity.

4. Join the event table extract(s) with the link table to producea traversal segment table.

5. Apply minimum criteria to the traversal segment table toproduce a list of segments that can be used for the trip.

6. Use the resulting table to create adjacent node entries foreach node record. These entries will allow the node table toserve as a turntable at intersections and to access points.Eliminate all other node table records or mark them as in-valid for this trip.

7. If desired, create an impedance factor for each traversal seg-ment based on such considerations as traffic volume, postedspeed limits, and/or turning movement volumes. This fac-tor can be used to select the most suitable path from amongthe alternatives available.

8. Use software functionality to find the “best” path based onstated criteria, such as the shortest distance or lowest imped-ance.

9. Consider the designation of universal traversals that will beused for all long-distance travel, such as rural interstate high-ways. This practice will simplify the pathfinding process bylimiting the work needed to finding a way onto and off ofthe long-distance facility. It will also help prevent the path-finding process from accidentally routing a vehicle onto aless desirable facility due to data errors and omissions.

The two sets of guidelines listed above are simplified; in prac-tice, the issues involved are a bit more complex. For example,most facility inventories are structured to use a single logicalcenterline for each roadway. The corresponding cartographicobject is a line down the middle of the road or in the median ofa divided highway. If the facility inventory stores data accordingto side of the road, then a mechanism that lets the user accom-modate this structure during dynamic segmentation will need tobe established. (We are not aware of any GIS software productthat handles this process automatically.) Below are a few optionsto address this data structure and to provide a cartographic ap-pearance that more closely matches feature geometry for dividedhighways:

1. Change the facility inventory to utilize a paved-course ortravel way centerline (i.e., a logical centerline for each sideof a divided roadway). This approach results in a set of threecenterlines, one for undivided roads, one for the right-sidepath, and one for the left-side path.

2. Construct the same line objects as would be needed forOption 1 but do not change the facility inventory. Instead,presort the inventory by side of road into three groups (left,right, and both sides).

3. Continue to use the logical centerline to connect to thefacility inventory but do not use it for map making. In-stead, use paved-course centerlines related to the logicalcenterline through a foreign key. This approach is the mostsatisfying option since it does not require any changes tothe facility database or preprocessing of the data; however,it does require some custom coding to transfer the resultsof dynamic segmentation to the correct adjacent paved-course centerline.

4. Continue to use the logical centerline to connect to the fa-cility inventory and write a custom routine that producesparallel lines for divided highways. This only works for small-scale maps where the difference between the position ofthe generated parallel lines and the true paths will not be aproblem.

Whatever option is chosen, the important starting point isto develop a set of business rules to guide the project. These busi-ness rules will include decisions about location referencing meth-ods. We continue to support the use of spatially based linearlocation referencing systems (linear LRSs) for linear transporta-tion features, such as highways, city streets, transit routes, andrailroads. Multidimensional and temporal LRSs should be over-laid on the spatially linear LRS for these features. However, theLRS is not the most important set of business rules to be devel-oped. That status is given to the rules governing transportationfeature identification. It does not matter how a feature and at-tributes are measured if the user cannot find the transportationfeature to which the attributes apply.

Transportation Feature IdentificationA transportation feature is an element of the transportation sys-tem that may be uniquely identified in the real world and forwhich attributes are provided (Butler and Dueker 1998). Sincethe central element of the multimodal Enterprise GIS-T DataModel is the transportation feature, there must be a way touniquely identify the feature across all modal components. Themodel defines transportation features by type within a jurisdic-tional framework that serves no other purpose than to provide acommon spatial context for expressing the location and extent ofthe facility. From an implementation perspective, more detailedspecification is required, one that accommodates both the needsof the database and the public user.

A relational database management system needs a uniqueidentifier for each record in a table. The design proposed below


uses a relational foundation consisting of tables (relations) con-nected to each other by foreign keys (relationships). In most cases,these identifiers are integer numbers that serve no other purposethan to uniquely identify a single record and they are difficult forusers to remember. For example, it is much easier to rememberthat a particular transportation feature is called SR 153 inHamilton County, Tennessee than that its primary key identifieris 286449. The identifier known to the general user is called apublic key. A form that we have suggested (Butler and Dueker1998) for the transportation feature identifier public key wouldproduce something like:

RDTN067ST00153000

whereRD is the transportation feature of type road;TN is the State of Tennessee as a United States Postal Ser-

vice (USPS) abbreviation;067 is the Federal Information Processing Standard (FIPS)

code for Hamilton County;ST is the designating authority, this illustrates a method for

identifying a state department of transportation (DOT);00153 is the road identification number; and000 is the sequence number for the section of 00153.

A working group under the auspices of the Federal Geo-graphic Data Committee’s Surface Transportation Subcommit-tee with support from the Bureau of Transportation Statistics hasdeveloped an alternative approach using many of the same con-cepts (FGDC 2000). This proposal is an extension to the Na-tional Spatial Data Infrastructure (NSDI) for uniquely identifyingFramework Transportation Segments. Our sample public keyexpressed as a transportation segment identifier using this pro-posed standard could be:

47001.S.001530000

where47 is the FIPS code for the State of Tennessee;001 is the designating authority of the state DOT (unde-

fined in the proposal but unique within the state);S is the segment type of transportation feature; and067153023 is the road identification number for section 23

of State Road 153 in Hamilton County (067). Thisnumber is unique among the designations made by theauthority for this feature type (undefined in the pro-posal but unique within the state).

Regardless of the format used, the public key should be sepa-rate from the internal database key because the public key maychange but the internal key must not—unless necessary to re-flect the replacement of one feature with another. Used in thisway, the primary mechanism for data sharing is reduced to sim-ply adding a new column in a data table to store the public key.

Once this equivalency has been established, future data exchangesbecome relational joins of an internal table with one provided byanother person. Cartographic conflation would not be used, al-though it may have been of assistance in the original equivalencydetermination.

Two points need to be emphasized here. First, the proposedidentification method is for data interchange or sharing and isnot necessarily for internal databases. Each agency providing orusing transportation data could continue to use identificationunique to the agency. The proposed public identifier is for “trans-lating” one identification system to another. For this system towork, each agency would need to establish a new field in theirfile to store the public key identifier, and then enter the appro-priate identifier for each transportation feature.

This brings up the second point. It is very likely that eachuser will have different transportation features. This means thatthere will not be a one-to-one correlation between the transpor-tation features defined using the public identifiers and thosepresent in user databases. These differences are accommodatedin both example methodologies listed above by the use of a three-digit segment sequence number. Differences can be accommo-dated as long as the segments comprise the same main feature.This means that all users will need to agree on transportationfeature endpoints but not on intermediate break points. A stateDOT may break transportation features at county lines, while anindividual county or city breaks them at every intersection. Theoverall transportation feature identifier would be the one thatends in “000.” As long as there are fewer than 999 segments, theproposed methodologies can accommodate any segmentationmechanism. Only at the highest level do the participants need toagree on transportation feature definition. Such agreements arerequired for any continuing data exchange process, so this is notreally an extra burden imposed by the proposed methodologies.

Linear Location Referencing SystemDatumBefore describing the database design, some guidance on estab-lishing a linear LRS datum is in order. The datum is the frame-work within which a location is specified. Transportation agenciesgenerally utilize an implicit linear LRS datum. The term “im-plicit” is used to mean that an origin and path are specified, butactual datum requirements and features for measurement accu-racy are not usually expressed. We contend that an explicit da-tum consisting of a network of anchor points (well-definedphysical locations) and anchor sections (the roadway segmentsbetween anchor points) must be established to provide the tem-poral and spatial measurement consistency required for reliabledata exchange. Anchor points have location references as manda-tory attributes. Anchor sections have direction and length asmandatory attributes. Such a datum design must provide a set ofrules for defining, selecting, and locating anchor points and an-chor sections, and for measuring the length of anchor sections.“Of particular concern are the identifiability and recoverability(persistence) of anchor points” (Vonderohe and Hepworth 1996,


p. 3). Simply put, transportation features and their attributescannot be left to float around, changing from database to data-base and from year to year—unless a change is necessary to re-flect a true change to the system.

While a linear LRS, such as route milelog, may locate eventson linear transportation features, a means is needed to locate thosefeatures in the real world and to define the measurement frame-work. To meet these needs, a system of anchor points and anchorsections must be established (Figure 1). The position of each an-chor point would be defined in all desired linear and nonlinearLRSs. Determining the location of anchor points is a good appli-cation of the Global Positioning System for GIS-T, since a Glo-bal Positioning System location for each anchor point can beused to tie the roadway to other coordinate systems. Anchor sec-tions, which extend from one anchor point to another along thepath of a transportation feature, have direction and length as theirprimary attributes. Anchor points and sections are geodetic ob-jects that establish the geographic datum of the linear LRS. An-chor section length serves as a quality control check for theaccuracy of linear LRS measurements. The anchor section fol-lows the logical centerline of the roadway, with anchor pointsbeing located along that centerline.

Anchor points may be difficult to precisely identify and cap-ture in the field, as they are usually located on an abstraction ofthe road (i.e., the roadway centerline). Anchor points need to betied to reference objects, which are physical locations that a usercan easily observe in the field and on maps (Vonderohe andHepworth 1998). Reference objects can be anything that is notreadily movable (e.g., a curb intersection, bridge end, traffic sig-nal pole, or survey marker).

To establish a linear LRS datum, anchor points must mini-mally be placed at the beginning and end of each roadway. Inter-mediate anchor points may be located along this base anchorsection. The density of anchor points is determined by the de-gree of positional accuracy desired; the greater the desired accu-

racy, the greater the number of anchor points needed (Vonderoheand Hepworth 1998). The ability to detect and correct measure-ment errors varies directly with the number of anchor points andinversely with the average length of anchor section.

To be valid, a datum must be tied to physical, real-worldlocations that are unambiguously defined. This would seem toeliminate such field references as county lines and other jurisdic-tional boundaries, since county line and city limit signs may notbe properly and/or consistently placed. However, the TennesseeDOT linear LRS, for example, requires the origin to be at thebeginning point of the road in the county. The reconciliation ofthese two needs is to tie the jurisdictional boundary to a refer-ence point that is unambiguously defined (i.e., make the loca-tion of the beginning anchor point and transportation featureorigin 0.000 at the jurisdictional boundary, but locate the bound-ary (and origin) as an offset from a reference point). The road-way is thus unambiguously tied to a datum-compatible location.

One or more anchor sections must be created to provide ageographic network reference for each roadway. Vonderohe andHepworth (1998) provide a way to calculate existing measure-ment accuracy and to determine the number of anchor pointsneeded to reach a given level of accuracy.

It is important to note the differences between accuracy, pre-cision, and resolution. “Resolution” is the closest proximity ofobjects that can be represented as being at different locations.For example, the Tennessee DOT roadway inventory uses a reso-lution of 0.01 mile (52.8 feet), which means that, to be placed atdifferent locations, two objects must be at least half that far apartfor measurement rounding. Thus, this linear LRS has a resolu-tion of 26.4 feet. “Accuracy” refers to the closeness with which aset of measurements approximates the true value, which cannotbe absolutely known. Anchor points and anchor sections providea way to reach a specified level of accuracy by controlling system-atic error; however, accuracy is ultimately determined by mea-surement methods. “Precision” refers to the repeatability of these

Figure 1. Roadway feature with anchor points, anchor sections, and reference objects.


Figure 2. The Transportation Feature Database physical data model.


methods. For high precision, all measurements for the locationof a feature are numerically similar. When precision is high, er-rors in measurement can be corrected through a uniform adjust-ment that compensates for the method’s systematic error. Whenprecision is low, errors cannot be readily corrected by adjustmentbecause random error is much larger than systematic error.

An increase in the density and location precision of anchorpoints results in an increase in linear LRS measurement accu-racy. The overall accuracy of the linear LRS is limited by theprecision of linear offset measurements from anchor points tolocations of interest along a transportation feature. Measurementsmade with distance-measuring instruments have errors that in-crease with distance (i.e., are said to propagate). Anchor pointaccuracy requirements should be determined by looking at boththe needs of users and the ability of available field procedures. Amultimodal linear LRS datum must meet the highest accuracy,precision, and resolution requirements of any modal databaseand application that it is intended to support. Applications re-quiring lower levels of spatial and temporal accuracy and preci-sion can utilize resolutions expressed with fewer degrees ofcertainty.

Transportation Feature DatabaseDesignThe primary transportation feature-attribute tables needed for acomplete multimodal transportation facility inventory form thecentral element of the proposal (Figure 2). The included trans-portation features in our sample database design are roadways,airport runways, waterways, railroads, and intersections. The pri-mary key of each table is underlined. The primary keys of datatables are designed to store history through the inclusion of atime stamp (entrydate and entrytime). A field is also provided torecord the name of the person who made the entry. The Trans-portation Feature Table includes a data item called the extkeyID,which is an external key identifier to link this table with an exter-nal data table. For instance, the extkeyID could contain thewaterwayID for linking a water-based transportation featurerecord to the Waterway Table. The value of the tranfeattypeIDwill determine the feature table to which the identifier in extkeyIDis related. This approach allows full normalization of the data-base using look-up tables and simplification of the naming pro-cessing. The included tables are described in the following sections.

Being centered on physical transportation features, the de-sign treats utilizing modes as events. For example, a transit routewould be a traversal across one or more transportation featuresegments (raillines and/or roadways), with each segment definedas a linear event on a transportation feature. A useful design fortraversals and other elements of a complete GIS-T database isprovided in our previous work (e.g., Dueker and Butler 1998).Time stamps are provided to support temporal applications, suchas the evaluation of traffic accidents based on the nature of theroadway network at the time of the events. “Person stamps,” inthe form of user identifiers, are provided as a managerial meansof tracking changes.

The sample tables illustrate the nature of attribute fieldsthat can be included. No intent should be inferred from theabsence of a particular attribute. Where justified by their po-tential usefulness, tables and fields have been included that of-fer benefits in implementing the design. For example, weincluded look-up tables for such defined domain variables asdesignator and direction. Not all tables need to be utilized; manyare included here to illustrate the multimodal flexibility of theproposal. Field names have been selected for their mnemonicvalue, but are not otherwise critical.

Transportation Feature TableThe Transportation Feature Table contains the data needed todescribe each feature in the transportation network. There willbe one record for each physical facility on the base map. Thetable uses tranfeatID, plus the date and time the record wascreated, as the primary key to identify each record. The de-scriptive data include the beginning and ending milepoints, astandard name, a separate external key (usually that of the datasource), and the direction of travel. The design assumes that allincluded transportation features will be of the linear type. Non-linear features may be referenced to adjacent linear features.For example, an airport terminal may be tied to a point on theaccessing roadway.

The table below provides additional details regarding eachincluded data item. In this example, the jurisdiction domain con-sists of counties in a single state, so a county line represents aforced end to each linear feature.

Data Item Meaning External Keyentrydate the date that the record was createdentrytime the time that the record was created

enteredby the user identification (ID)

of the person creating the recordtranfeatID the unique numeric identifier

for a transportation featuretranfeattypeID the unique identifier for a

transportation feature type YesstateID the unique identifier for the

record containing theUSPS state code Yes

countyID the unique identifier for therecord containing theFIPS county code Yes

designatorID the unique identifier of the typeof agency defining the feature Yes

primaryID the unique primary route identifiercreated by the designator Yes(all related features will carrythis same primary identifier)


Data Item Meaning External KeysecondaryID the unique secondary route

identifier created bythe designator Yes(realignments, ramps, andservice roads will carry asecondary identifier otherthan “000”, which indicatesthe existing mainline)

beginmilelog the milelog measure for thefeature’s origin

endmilelog the milelog measure for thefeature’s terminus

directionID the unique identifier of adirection code

The fields tranfeattypeID, stateID, countyID, designatorID,primaryID, and secondaryID would be combined to create a singlepublic key for accessing the data without knowledge of the inter-nal key (tranfeatID). This public key conforms to our previouslypublished naming convention proposal (Butler and Dueker 1998).Other forms for this key are supported by the table’s design.

The public key fields could be combined into a single field,but the grouping of selections, such as selecting all the roads in agiven county, would be more difficult due to the need to parsethe combined value. External attribute tables, such as the EventTable in Figure 2, utilize the internal primary key of the Trans-portation Feature Table as the necessary foreign key connection.To be placed in geographic order, records could be sorted by ei-ther the internal or external key and beginmilelog.

It is important to note that the actual key field values are notstored here. They are kept in look-up tables, the population ofwhich defines the domain for each public key component. Theunique record identifier for the applicable public key compo-nent value is stored in the Transportation Feature Table. Thisdesign concept allows domain value meanings to change overtime without requiring changes in the Transportation FeatureTable. For example, if the decision is made at some point to uti-lize the two-digit FIPS state code instead of the initial USPS ab-breviation, only the State Code Table needs to be updated,assuming that the identifier for each state (stateID) remains un-changed. It would be a simple matter to concatenate and expandthe stored identifiers to produce a single public key field in a dataextract for exchange with another user.

The public key fields do not form the internal primary keyof the table, which is simply tranfeattypeID plus the time stampfields (entrydate and entrytime). The time-stamp fields providethe option to store changing transportation feature descriptionsover time. The concept could be extended to include two similarfields that stored the valid time period for the record (i.e., from abeginning date to an ending date). These values are independentfrom the time stamp shown, which is intended to store the timethat the record was last updated.

Transportation Feature Type TableThe type of transportation feature being described in a Transporta-tion Feature Table record is determined by the value of tranfeattypeID.The Transportation Feature Type Table stores the domain for thisfield and serves as a mechanism for generating a pick list or pull-down menu of choices for data entry and reporting. To support thisfunction, a listorder field is supplied that allows a database adminis-trator to sort records in this table for presentation to the user. Thisfield is alphanumeric to allow alphabetic sorting. This approach canbe used to enter a value of “03h” to place an entry between “03” and“04,” so that inserted options do not require renumbering of allsubsequently listed choices. The field can also be used to change theordering of values over time (e.g., to make the more frequently usedvalues appear at the top of the list).

Data Item MeaningtranfeattypeID the unique numeric identifier for a

tranfeattypetranfeattype the code value representing the transporta-

tion feature type (e.g., RD or AR)tranfeattypeTXT the longer description of type (e.g., road or

airport runway)listorder the alphanumeric indicator of list display

order

Only the pick lists relevant to the concepts expressed by thisillustrative design—that a single database can be constructed fora multimodal agency’s GIS-T infrastructure—are described inthis article. Look-up tables implied by the use of such field namesas ststatusID and dirprefixID in the Street Table are not included,as the examples listed are sufficient to express the concept.

Anchor Point TableAnchor points define the locations on the roadway system that canbe readily identified on maps as an aid to conflation, the process ofcombining two databases (maps) to create a new one. Anchor pointsalso provide the means to define roadway segment locations usinga two- or three-dimension coordinate system, such as a State PlaneCoordinate System, as a supplement to the linear LRS measuresthat may be resident in an inventory database.

Anchor point locations are stored in this table as a text de-scription, such as “the intersection of Maple Street and BroadAvenue.” The physical location of this intersection is defined byone or more reference objects and the bearing and distance fromthe object(s) to the anchor point. The location of each anchorpoint in the linear LRS of roadways is specific to the roadway, asdefined by one or more anchor sections. Thus, the linear LRSlocation of an anchor point is part of the Anchor Section Table.

Data Item MeaninganchorpointID the unique numeric identifier for an an-

chor pointentrydate the date that the record was createdentrytime the time that the record was created


enteredby the user ID of the person creating therecord

descTXT the description of the anchor point location

Fields could be easily included to store other items of inter-est such as anchor point type (intersection, bridge end, countyline, etc.) and the date that the anchor point was established.

Anchor Section TableAn anchor section begins and ends at anchor points. Each an-chor section provides a highly accurate field measure (length) forthe roadway segment. The number of anchor points and anchorsections are determined by the level of field measurement accu-racy required. Anchor section records provide the linear LRSmilelog measure for each terminating anchor point, the distancealong the measured route between them, and the date and bywhom the length measurement was made. One could extend thelist of anchor section attributes to include how the measure wasmade and similar descriptive items.

Data Item MeaninganchorsectionID the unique numeric identifier for an an-

chor sectionentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordroadwayID the unique numerical identifier for a road-

waybeginanchorpt the identifier of the anchor point at the

section originbeginmilelog the milelog measure for the anchor section

originendanchorpt the identifier of the anchor point at the

section terminusendmilelog the milelog measure for the anchor section

terminuslength the anchor section length, in units of 0.001

milemeasuredate the date that the length measure was takenmeasuredby the name of the person taking the length

measure

Reference Object TableOne or more reference objects may be used to locate each anchorsection. The initial base map is not expected to contain referenceobjects, but the data table needs to be created as part of the origi-nal database so that appropriate records may be added over time.Reference objects may be of various types, such as survey monu-ments or iron pins. A given reference object may be defined inmore than one datum or measurement system.

Data Item MeaningrefobjectID the unique numeric identifier for a refer-

ence objectdatumID the datum in which the reference object

location is definedentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordrefobjtypeID the unique numeric identifier for the refer-

ence object typedescTXT the text description of the reference object

and its locationxcoord the x-plane coordinate measure for the ref-

erence object locationycoord the y-plane coordinate measure for the ref-

erence object locationzcoord the z-plane (height) coordinate measure for

the reference object locationmeasuredate the date that the location measure was

takenmeasuredby the name of the person taking the location

measure

Reference Object Type TableReference objects may be of several types, each representing akind of real-world object that may be used as a reference object(e.g., iron pins, concrete monuments, and traffic signal poles).The listorder variable is alphanumeric and can be used to sortitems on a pull-down pick list.

Data Item MeaningrefobjtypeID the unique numeric identifier for a refer-

ence object typerefobjtype the code value representing the reference

object type (e.g., LP)refobjtypeTXT the type code meaning (e.g., light pole)listorder the alphanumeric indicator of the list dis-

play order

Reference Object/Anchor Point TableA single reference object may be used to locate multiple anchorpoints, or one anchor point may be positioned using multiplereference objects. The resulting many-to-many relationship re-quires a resolution or associative table. To reflect expected fieldsurvey methods, this table also stores the bearing and distancefrom the reference object to the related anchor point, instead ofx,y measures.

Data Item MeaninganchorpointID the unique numeric identifier for an an-

chor sectionrefobjectID the unique numeric identifier for a refer-

ence object


apdistance the distance in units of 0.001 foot to theanchor point

apbearing the compass bearing to the anchor pointusing the reference object as the pivot pointand true north as the zero point of rotation(see Figure 1)

Event TableEvents are the characteristics and subordinate features of alinear transportation feature. Characteristics include suchthings as the functional class, number of lanes, pavement type,speed limit, and shoulder type. Subordinate features includeintersections, bridges, signs, and guardrails. The example EventTable design accommodates temporal references with a dateand time stamp. It accommodates divided highway attribu-tion through a side field; a similar approach could be used tohandle data by lane. Potential values for side are left, right,and both. The data model shows anchor points and anchorsections being described as events. Reference objects could bestored as events using the lateraloffset and side fields to showa distance and direction from the transportation feature tothe object. Point and linear events have a beginning milelogmeasure, while only linear events have an ending milelog mea-sure. The eventID field is proposed to be unique only withinthe transportation feature; although it could be universallyunique across the entire database, this is not required. Thevalue of tranfeatID is a partial foreign key to connect the eventto its own transportation feature. In application, software couldsearch for the most recent Transportation Feature Table recordor one consistent with the time stamp in the Event Tablerecord. As with the Transportation Feature Table, additionalfields could be incorporated into the design to store the pe-riod of time for which the record was applicable.

Data Item MeaningtranfeatID the unique numeric identifier for a trans-

portation featureeventID the unique numeric identifier for an evententrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordeventtypeID the identifier for an event typebeginmilelog the milelog measure for the event originendmilelog the milelog measure for the event terminus;

[null] for point eventsside the side-of-road indicator (left, right, or

both)lateraloffset the distance from a logical centerline to an

object located off the roadwayvalue the value of the specified attribute

Event Type TableEvents may be of several types (e.g., functional class, number oflanes, pavement type, speed limit, shoulder type, intersection,bridge, sign, or guardrail). In the example database design, eachevent type has its own parameters for units of measure and num-ber of decimal places. The number of decimal places is neededbecause the value field in the Event Table is alphanumeric; there-fore, it cannot be stored in a numeric format that includes deci-mal information. This means that a data extraction process willneed to use the numdecimal value for the event type to create theproper appearance. An alternative is to store an explicit decimalpoint in the value field.

Data Item MeaningeventtypeID the unique numeric identifier for an event

typeeventtype the code value representing the event type

(e.g., post-mounted sign)eventtypeTXT the type code meaning (e.g., post-mounted

sign stored as MUTCD type)unitofmeasure the units of measure for the event typenumdecimal the number of decimal places in the field, if

numericlistorder the alphanumeric indicator of list display

order

State TableThe U.S. Postal Service has established a set of state abbrevia-tions that we have endorsed as a geographic naming standard forthe transportation feature public key. This look-up table storesthe relevant values. An alternative would be to use the FIPS statecodes, as proposed by the current NSDI initiative.

Data Item MeaningstateID the unique numeric identifier for a statestateabb the abbreviation for the state (e.g., TN)stateTXT the name of the state (e.g., Tennessee)listorder the alphanumeric indicator of the list display

order

County TableThe FIPS catalog includes a standard on numeric county desig-nations for use in computer applications. This look-up table storesthe relevant values. A multi-state agency might include the statecomponent of the FIPS county code. States have a two-digit des-ignation, and the county designations are three digits in length.

Data Item MeaningcountyID the unique numeric identifier for a FIPS

county codecountyFIPS the code value representing the county

(e.g., 065 or 47065)


countyTXT the name of the county (e.g., HamiltonCounty, TN)

listorder the alphanumeric indicator of the list dis-play order

Direction TableThe Direction Table stores the domain values for the direction oflinear measurement on transportation features. Linear LRS mea-sures generally follow the federal standard of direction (i.e., westto east and south to north). Other approaches may be used, in-cluding bi-directional methods for divided highways.

Data Item MeaningdirectionID the unique numeric identifier for a com-

pass direction of traveldirection the code value representing the direction

(e.g., 3 or SB)directionTXT the full name of the direction (e.g., south-

bound)listorder the alphanumeric indicator of the list dis-

play order

Designator TableThe naming convention proposed by us in an earlier article andthe subsequently proposed NSDI transportation feature identi-fication process include a component for the designating agency.Using such a component in the public key precludes the need fora universal agent that ensures unique values for identifiers. (Thereis, of course, the issue of making sure that designator codes areunique, a subject beyond the scope of this article.) With the des-ignating agency component, these agencies may act independentlyto ensure that the transportation feature identifiers are globallyunique. Standards can be developed for the construction of anagency designation code such that they are automatically unique,such as by using their federal tax identifier.

Data Item MeaningdesignatorID the unique numeric identifier for a desig-

nator of transportation feature identifiersdesignator the code value representing the designator

(e.g., 074)designatorTXT the name of the designator (e.g., Hamilton

County TN Highway Department)listorder the alphanumeric indicator of the list dis-

play order

Airport Runway TableThe database design in Figure 2 includes five transportation fea-ture types. These tables extend the database to include informa-tion retained by individual modal offices in a multimodal agency.In some cases, such as that of the Airport Runway Table, nochanges can be made in the legacy modal table to point back tothe corresponding Transportation Feature Table record’s identi-

fier. Instead, one would be required to use runwayID in the Air-port Runway Table as a foreign key to select the correct record inthe Transportation Feature Table using extkeyID andtranfeattypeID.

Data Item MeaningrunwayID the unique numeric identifier for an airport

runwayentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordairportID the identifier for an airportrunway the compass direction pair for the runway

name (e.g., 090-270)descTXT the long description of the facility (e.g.,

main east-west runway)pavetypeID the identifier for the runway pavement

type

Rail Line TableThe Rail Line Table stores information about railroad tracks,potentially including light rail and other public transit forms.Many railroads use linear LRS measures, so this table could beexpanded to include them. Another option would be to includeterminating facilities for each rail segment, such as a switch orblock signal.

Data Item MeaningrailID the unique numeric identifier for a railroad

trackentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordrrcoID the identifier for a railroad companyraillinebranch the railroad service area (e.g., Chesapeake

Subdivision)descTXT the long description of the facility (e.g.,

Mainline from Signal 342 to Switch 9477)

Although this article does not fully explore the application,the concepts could be equally applied to public transit on sharedand dedicated roadways. For example, bus routes could be ex-pressed as linear events and/or as transportation features in theirown right. Route segment travel time would be a useful attribute.

Waterway TableWaterways are transportation facilities and impediments to sur-face travel; they may be present in a multimodal transportationdatabase for either reason. The Waterway Table could serve as alook-up table for selecting a feature crossed by a structure in abridge inventory.


Data Item MeaningwaterwayID the unique numeric identifier for a water-

way or water bodyentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordwatername the name of a waterway or water body

(e.g., Tennessee River)waterabb the short name to put on maps (e.g., Ten-

nessee R.)mainstreamID the identifier for the waterway into which

this feature flows (e.g., Ohio River)descTXT the long description of the facility (e.g.,

Tennessee River within ChickamaugaLake)

watertypeID the identifier for the water type

Street TableUnlike the Airport Runway Table, the Street Table has been de-signed to include tranfeatID as a direct foreign key to point tothe related Transportation Feature Table record. We have taken adecidedly local government approach in showing a Street Table,using address numbers as a linear LRM. The four street namecomponent fields (dirprefixID, name, stypesuffixID, anddirsuffixID) follow the standard endorsed by the National Emer-gency Number Association for E-911 messages. The dirprefixID,dirsuffixID, and adddirectionID fields could point to a singledirectionID field in the Direction Table. A subordinate BlockTable could also be included.

Data Item MeaningstreetID the unique numeric identifier for a named

streetentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordtranfeatID the identifier for the transportation featureminaddress the minimum address number for the

streetmaxaddress the maximum address number for the

streetststatusID the identifier for the street status (to indi-

cate open, closed, planned, etc.)dirprefixID the identifier of the street name direction

prefix, if anyname the root street name (could include street

type prefixes common in names derivedfrom foreign languages)

typesuffixID the identifier of the applicable street typedirsuffixID the identifier of the direction suffix, if anyadddirectionID the identifier of the direction of increasing

address numbers

Intersection TableWe propose that intersections have a separate table. This designapproach provides a simple means for storing information re-garding intersections, such as traffic control and turn restrictions,in a normalized database design. A normalized design precludesthe need for redundant storage of intersection information forall intersecting street segments. The design also provides a “poorman’s” topology for pathfinding. Using the intersection type field,the meaning of intersection can be extended to include any typeof junction, from limited-access highway interchanges to bridgesand railroad grade crossings. The intersection identifier is pro-posed to be universally unique. Since an intersection is also anevent on the two or more involved linear transportation features,the various many-to-many relationships of intersections and trans-portation features would be stored in the Transportation Fea-ture/Event/Intersection Table, a resolution or associative table.

Data Item MeaningintersectionID the unique numeric identifier for an inter-

sectionentrydate the date that the record was createdentrytime the time that the record was createdenteredby the user ID of the person creating the

recordinttypeID the identifier for the intersection type

Transportation Feature/Event/Intersection TableThis table is an associative or resolution table that stores the many-to-many relationship between intersections and the events thattie them to transportation features. The table presented here con-sists only of foreign keys; however, they can include attributes.We have chosen to include the complete key of the Event Tablebut only a partial key for Intersection and Transportation Fea-ture table records. This choice means that query software willneed to pick the Transportation Feature and Intersection tablerecords that best matches the timing of the Event Table record.An obvious alternative is to include the complete primary key ofall three tables.

Data Item MeaningintersectionID the identifier for an intersectiontranfeatID the identifier for a transportation featureeventID the identifier for an evententrydate the date event that the record was createdentrytime the time event that the record was created

ConclusionThis article demonstrates an explicit path to meet the primaryrequirements for a multimodal transportation database that canserve the needs of an enterprise GIS-T infrastructure. Relatedguidance can be derived from articles noted in the list of refer-ences. Our intent is to show that it is possible to construct suchan infrastructure regardless of the current state of informationsystems. Whether it is GIS- or computer-aided design-based, any


successful enterprise solution will need to adopt an “all of theabove” approach that includes all existing and potential applica-tions. The key is not in the cartography, programming languageor software platform currently used by various offices, but in thedesign of the database that connects them to each other.

The proposed Enterprise GIS-T Data Model has been shownto include the various data models now used by the many func-tional units of a multimodal transportation agency (Butler andDueker 2000). For the sake of simplicity, only the elements ofthe model that serves linear transportation facilities are includedhere. Previous articles by the authors describe a more robust modelthat includes the area and point features, includingnontransportation facilities.

The Enterprise GIS-T Data Model supports isolated dataexchanges of component pieces, such as cartography or featurecharacteristics, by unbundling the elements of the presently domi-nant integrated topological vector data model. Relatively simpleand inexpensive software currently exists to meet the majority oftransportation agency needs. The Enterprise GIS-T Data Modelisolates the data from software technology by allowing data-cen-tric systems to be constructed in all currently available productlines that support dynamic segmentation. This article demon-strates one way in which the model can be implemented.

About the Authors

J. Allison (Al) Butler is Director of GIS for Hamilton County,Tennessee, where he leads a growing community GIS enter-prise. Previously, Butler was with the Florida DOT, and heserves as a part-time consultant to other state and local trans-portation agencies. He is certified by the AICP and theASPRS, and earned a B.B.A. from the University of Georgia(1974). The author may be contacted at Hamilton County,Tennessee, 117 E. 7th St., Suite 300, Chattanooga, TN37402. [email protected].

Kenneth J. Dueker is the Director of the Transportation StudiesCenter at Portland State University, where he previouslyserved as the Director of the Center for Urban Studies. Dr.Dueker holds a B.S. from the University of Washington(1960), an M.S. from the University of Washington (1963),

and a Ph.D. from the University of Washington (1967). Hisresearch interests include GIS applications and transporta-tion planning. The author may be reached at the Center forUrban Studies at Portland State University, Portland, OR97207-0751. [email protected].

References

Butler, J.A. and K. Dueker, 2000, A Primer on GIS-T Databases.Portland, OR: Center for Urban Studies, Portland StateUniversity. http://www.upa.pdx.edu/CUS/PUBS/contents.html

Butler, J.A. and K. Dueker, 1998, A Proposed Method of Transpor-tation Feature Identification. Portland, OR: Center for Ur-ban Studies, Portland State University. http://www.upa.pdx.edu/CUS/PUBS/contents.html

Dueker, K. and J.A. Butler, 1999, A Framework for GIS-T DataSharing. Portland, OR: Center for Urban Studies, PortlandState University. http://www.upa.pdx.edu/CUS/PUBS/contents.html

Dueker, K. and J.A. Butler, 1998, GIS-T Enterprise Data Modelwith Suggested Implementation Choices. Journal of the Ur-ban and Regional Information Systems Association, 10 (1), 12-36.

Federal Geographic Data Committee, Ground TransportationSubcommittee, 2000, NSDI Framework Transportation Iden-tification Standard—Draft Number 3. Washington, DC: USDOT Bureau of Transportation Statistics. http://www.bts.gov/gis/fgdc/web_intr.html

Opiela, K.S.,1997 Research Results Digest No. 218: A Generic DataModel for Linear Referencing Systems. Washington, DC: Trans-portation Research Board.

Vonderohe, A., and T. Hepworth, 1998, A Methodology forDesign of Measurement Systems for Linear Referencing. Jour-nal of the Urban and Regional Information Systems Associa-tion, 10 (1), 48-56.

Tell Someone About URISA Today!

URISA Journal � Harvey 29

Collaboration And Socio-technicalConstructionInformation technology researchers usually examine collabora-tion in the context of specific tasks or functions. Normally pre-sented from a system engineering approach, the implementationof geographic information systems (GISs) considers isolated func-tional and technical issues. This approach extricates organiza-tional and institutional issues and emphasizes tasks in a Tayloristanalysis of functionality. Literature on GIS diffusion emphasizesthe organizational setting in which agencies develop GIS in rela-tion to administrative activities and GIS functions. Diffusion lit-erature (Rogers 1983) contextualizes functions in theirorganizational setting.

Previous organizational research has shown the importanceof social factors for GIS (Calkins and Obermeyer 1991, Nedovic-Budic and Godschalk 1994, Pinto and Azad 1994, Campbelland Masser 1995). Designing and developing GIS is a complexprocess involving a diverse range of professionals and adminis-trators, among other groups. Research on data sharing demon-strates that institutional factors are the greatest impairment indeveloping GIS (Onsrud and Rushton 1995). Literature on dif-fusion underscores the importance of institutional factors.Campbell and Masser show that the effectiveness of GIS in prac-tice is directly tied to the institutional adoption of the technol-ogy. Working in the diffusion paradigm, Assimakopoulos (1997)traced the social network in the GIS community in Greece andexamined how GIS innovations were diffused in this network.

In traditional sociology, a network consists of individualnodes that exchange with each other through conduits for com-munication. The classical sociological understanding of a net-work focuses on relationships between individuals asrepresentatives of groups. This work complements the actor net-work theoretical framework deployed in this paper to understand

Constructing GIS: Actor Networks of Collaboration

Francis Harvey

Abstract: The social coordination of geographic information technologies relies on collaboration between actors from the public,private, and education sectors. Diffusion, implementation, information sharing, and studies of the use of geographic informa-tion systems (GISs) examine the collaboration in relation to specific activities. Applying concepts from actor network theories, thisarticle examines the socio-technical context. The construction metaphor distinguishes social network approaches from actornetwork approaches. This research provides insight into the relationships of GIS socio-technical networks, which are invaluablefor understanding the alliances, data sharing arrangements, and standards necessary for specific GIS tasks or functions. Theresults of research suggest that collaboration involve the construction and maintenance of hybrid networks that connect multiplehuman and non-human actors into strategic alliances. An important finding is that technologies are among the “key players” inthe GIS community. Standards, an organizational technology, focus strategic alliances and involve diverse groups in mutuallybeneficial projects. These groups are in long-term relationships that the introduction of GIS technologies can substantially alter.

the context in which social coordination of GIS occurs. Wheresocial networks study specific interactions between individualsand groups, actor network theories focus on examining actors’wide-ranging relationships and tracing activities in networks. Thesocial coordination involved in developing and using a GIS canbe understood through a construction metaphor and the col-laboration necessary to construct a building. Where implemen-tation and use studies examine construction in terms of the tasksinvolved, the construction metaphor looks at how activities arecoordinated.

A construction metaphor is also useful for distinguishingactor network studies from social networks. From the function-alist viewpoint that characterizes most social network studies,construction is an undertaking defined by an objective and real-ized through the coordination of many individuals and groups.It orients itself to a plan and follows it closely. In actor networkstudies, the emphasis is on the skills and expert knowledge of theworkers, craftspeople, and artisans, and their many relationships.In practical work, they must resolve unexpected challenges andmaximize opportunities through solutions that ensure the stabil-ity of the construction and augment their position in society. Inaddition to acquiring and maintaining skills, craftspeople andartisans must collaborate with persons from other trades. Theyare engaged in multiple networks that go beyond the construc-tion project at hand. Systems engineering and diffusion litera-ture consider contingent networks of relationships only withregard to a particular activity. The study presented here exploresthe nature of relationships without limiting the analysis to a par-ticular issue.

An additional strength of the actor network approach is thatit transcends the limits of the construction metaphor. While theconstruction of a physical structure is finished at a point, a GIShas interest by many in the community; it is usually never com-


pleted, but is always being “built.” In other words, GIS calls forconstant construction to meet the changing demands of its “dwell-ers.” This poses special challenges for studies of GIS construc-tion, as with other large technological systems. Actor networktheories provide a framework for understanding the ongoing co-ordination that precedes every building activity.

The activities of GIS technology specialists and profession-als are social and technical. A relevant characteristic for actornetwork studies is the conviction that specialists are not “ivorytower” dwellers, but are actually deeply mired in the practicalcontingencies of everyday practices. Construction and mainte-nance are messy. Callon (1987) and others suggest that special-ists such as engineers are not full-time technologists but arepart-time technologists and part-time sociologists. They are con-tinually enmeshed in relationships both within and outside theirdisciplines.

These points are especially pertinent for GIS. GIS is a“young” field that lies between many disciplines. Cartography,surveying, and computer science are pre-eminent, but many otherfields use and develop GIS. Such specialists as surveyors, plan-ners, and sanitary engineers come from distinct disciplinary back-grounds that involve constant arbitration with other groups. Therelationships between different groups cannot be reduced to onenetwork but consist of multiple networks with different degreesof membership. Just as people are not one-dimensional stereo-types, the subjects of this research carry out their activities andparticipate in networks in many different ways. For example, asurveyor may be a consultant most of the time, but will also needto program software, manage projects, converse with prospectiveclients, etc. Persons working with GIS wear different “hats” de-pending on the groups with which they associate. In actor net-work theories, these hats are different technological artifacts.Nevertheless, are these hats (software and hardware) just protec-tive devices and enhancements to personal safety? Michel Callon,Bruno Latour, and others associated with actor network theoriespoint out that they also indicate convictions and disciplinary af-filiations. By stepping back from a focus on distinct issues to aperspective that engages the activities of actors involved in differ-ent groups, researchers can develop a better understanding of thecontexts for the development and use of geographic informationtechnologies. This is the foundation for analyzing how existingnetworks and relationships affect the construction and use of tech-nologies. Actor network theories open the ways for GIS studiesto develop a broader understanding of the geographic informa-tion technology construction.

Collaboration as the NetworkThe network metaphor offers social scientists a powerful per-spective on the processes of social organization and coordina-tion. Social organization and coordination are key to successfulcollaboration. In previous studies of GIS networks, the work ofakopoulos (1997) stands out for its depth and detailed descrip-tion of the Greek GIS social network. As outlined above, actornetwork theories deploy a different understanding of networks.

This section presents a description of the actor network theoreti-cal framework used in this research.

Fundamental differences in actor network versus classic so-ciological networks are relevant to this research. Traditional so-ciological network theories are structuralist (Giddens 1979) andfind explanations for behavior in the tension between the poten-tial for action and a framework that leaves possibilities open, butdelineates possible action. Actor network theories are post-struc-turalist and hold that structures are also malleable and definedby action. In actor networks, humans and non-humans are bothnodes to be explained; the network “links” between the nodes aretraces of exchange (Mol and Law 1994). We can summarize anessential difference in what each type of network seeks to ex-plain. An explanation of structuralist approaches focuses on theanalytical contrast between structure and agency and the contin-gency of individual or group action. In actor network approaches,the methodological maxim is to “follow the actors” (Latour andWoolgar 1979) and explain what they do concerning other hu-man and non-human actors. Bear in mind that the network is illsuited to metaphorically suggest the emphasis of actor networktheories; it is a carryover from an earlier actor network theoryincarnation that was structuralist (Callon et al 1986).

Works by Latour (1979, 1987, 1993, 1999), Callon (1986,1991), and others (Akrich 1992, Demeritt 1996) on the actornetwork theory show that the network is a construct that indi-cates the relationships between actors. There are no links in thenetwork theory, only traces of relationships. This theoretical ap-proach is characterized in three ways: First, actor network theo-ries emphasize the dynamic character of the social exchanges andpolitical processes inscribing many interested parties in the build-ing and use of technologies. Second, as the network in actor net-work theories refers to the dynamic character of relationshipsbetween groups and individuals, it includes non-humans. Tech-nological artifacts are much more than surrogates for certain hu-mans; they are actors who bundle multiple intentions and act inways that complement and extend humans (Latour 1992). Com-mon technologies, such as the answering machine, have majorimpacts on interactions and the organization of groups. Third,the construction of artifacts is a process of developing coherencebetween multiple actors. Continuing the example, answeringmachines ease new relationships and unexpected behaviors, suchas conversation without the simultaneous presence of two par-ties. Technologies incorporate and merge different interests inbundled socio-technical relationships. In summary, the networkmodel for actor network theories is that nodes are people, insti-tutions, and artifacts; connections are agreements and exchanges.

The working definition of actor networks in this paper is:The traces of relationships between people, institutions, and artifactsconnected by agreements and exchanges.

A refined description of actor network stresses that this theo-retical framework explains the interactions without the strongreliance on contingency characteristic of Giddens’ structurationtheory and Bhaskar’s critical realist structuration theory. Actornetwork theories are as much ontologies as epistemologies (Mol


and Law 1994), in which coordination is an issue of relation-ships.

A post-structuralist explanation puts the burden on explain-ing how actors came to collaborate. There is no a priori commonground in actor networks. Institutions and technologies are notnatural but are socially constructed. Any common ground be-tween actors is the result of exchanges and involvement. Differ-ent interests must be translated (i.e., coherent standpointsdeveloped) and people from different groups enrolled. With anincreasing enrollment of actors, that position becomes strongerand more relevant to other groups. The ascendancy of a softwarepackage to a de facto standard is a simple but illuminating ex-ample. Early on, it may be just one of many packages. As largeagencies begin to use a particular package for characteristics thatsuccessfully distinguish it from other software packages, theseadvantages become inherent. Widespread use of the software fol-lows and it becomes a necessity for participating in the market.The control of the operating software market by Microsoft andthe proliferation of Windows software products is a case in point.

In actor network approaches, the network of relationshipsrepresents collaboration between humans and non-humans. Anetwork is constantly in flux; it is not a static infrastructure ofconnections that people rely on to do business or politics. Infor-mation technology development never stops (Schuler andNamioka 1993) and the networks are constantly changing.

Exploring Socio-technical Relation-shipsThis section presents an explorative research project that pre-cedes the examination of GIS actor networks. The results of thisresearch provided snapshots of issues. A survey was undertakento help guide in the formulation of interview questions. The find-ings of this study are presented only to contextualize the inter-view methodology. The initial survey sample was selected fromthe mailing list of the department of rural surveying at the SwissFederal Institute of Technology in Lausanne, Switzerland (EPFL).The survey was sent by mail to 632 persons who had been addedover several years to the mailing list. Two weeks after the originalmailing, a “reminder” was sent to all individuals who had notreturned their survey by that point. After 6 weeks, partially com-pleted surveys had been returned by 212 people. We evaluatedand analyzed the 128 fully completed surveys.The questions of the exploratory survey were divided into threesections:

1) Who are you?2) What do you do?3) Whom do you work with?

Question 1 was designed to gather information about eachrespondent’s relevant demographic characteristics, the type andlength of employment, education, and an evaluation of their ownGIS competence. Question 2 asks for information about therespondent’s work in more detail. Example queries are: How of-

ten does the respondent work with GIS, what kind of problemsdo they find are more important, and the number of GIS soft-ware packages they use. In Question 3, the responses are designedto determine the amount of time the respondents work with per-sons from different sectors and organizations and the number ofcontacts they maintain.

The research examines GIS activities occurring in commu-nities in the French-speaking portion of Switzerland. This in-cludes the cantons of Fribourg, Geneva, Neuchatel, Jura, Vallis,and Vaud. The population of these cantons is approximately 1.5M people. The results of the survey aided in the formulation ofresearch questions for interviews with persons from public, pri-vate, and education sectors in this area. Included were interviewswith representatives from these groups.

Who Works with GIS?All of the 212 respondents to the demographic questions are overthe age of 23 years and most are in their late 30s or early 40s.They are predominantly male. Only 6% of the respondents arewomen. A total of 37% of the respondents work in surveying-related fields, 40% work in private offices, and 60% work inorganizations employing fewer than 30 people.

Generally, the respondents have been doing the same activ-ity and working at the same job for a long time. A total of 43%have been carrying out the same activity for more than 12 yearsand 32% have been working at the same position for more than12 years. This cuts across sectors. In many ways homogeneous,the respondents are well distributed over the three sectors: pri-vate offices (40%), public administration (39%), and higher edu-cation (11%); 10% are employed in other activities. Few of therespondents have a doctorate (10%), and the majority have alicense, master’s degree, or professional diploma (74%). Interest-ingly, although 76% consider further training and seminars im-portant, only 44% of the respondents indicate that they havehad specialized GIS training.

What Do GIS Users Do?The respondents to the questionnaire work with a diverse set ofpersons within a wide range of organizations. Fewer than 28% ofthe survey respondents work daily with GIS, and only 14% con-sider themselves to be specialists (see Tables 1 and 2). The tasksthat they work on range from landscape architecture to survey-ing. The actual use of GIS in collaborative work with personsfrom the same and different sectors was raised in the last sectionof the interview (see the following section) and in interviews whenspecific tasks could be explored in terms of collaboration.

The majority of respondents work for private organizations(40%). They also work for small organizations with small GISgroups. Thirty percent of the respondents work for organizationswith fewer than 10 employees, and 30% for organizations withbetween 11 and 30 employees. The number of employees work-ing specifically with GIS is also small. A total of 78% of therespondents indicated that fewer than 10 persons in their organi-zation work with GIS. Curiously, in response to the question of


whether the respondent works more with technical or organiza-tional problems, 41% say they work mostly on technical prob-lems and 23% work usually on organizational problems; however,36% show that they their work is both technical and organiza-tional. This response suggests the importance that respondentsafford dealing with both technical and organizational problems.

Interestingly, although 4% of the respondents never workwith GIS directly and do not consider themselves GIS-compe-tent, they do work on GIS-related issues. Although almost 30%work on a daily basis with GIS, the majority of respondents workless frequently with GIS. The small size of the organization andthe many activities described by the respondents suggest that GISis one component of their work; only in the rarest cases is GISthe sole or main focus of work. Again, the majority of respon-dents do not consider themselves specialists for GIS. More thanhalf of the respondents to the survey indicate that organizationalactivities are more or as important as technical issues.

In other words, GIS is but one of the tools used for diverseactivities. The workload of respondents involves many more tasksthan GIS alone. The role of GIS-related tasks in their work couldvary widely.

With Whom do Respondents work?Instead of requesting names of collaborators in the questionnaire,which would supply an exact description of individual activitiesin the networks but be met as a possible invasion of privacy orbusiness interests, general questions regarding collaboration of-fer insight into disciplinary relationships and involvement in ac-tivities. This format preserves the anonymity of the respondents.

Responses to questions about actual contact time with otherindividuals inside and outside of the organization provided is-sues to pursue in interviews. Although on average, 47% of therespondents’ time is spent with contacts from public administra-tions, the contact time with municipalities, cantons (states), andfederal organization varies greatly by organization type. The mostcontacts any individual in the study has to cantonal organiza-tions are 30, but the average is only two. This suggests that asmall number of people have a very broad network. Most peoplehave smaller networks.

Most surprising is the low importance that the respondentsgive standards. Thirty percent indicate standards are of no im-portance, 26% indicate that they are of little importance, 21%say they are important, and only 22% say they are very impor-tant or essential. In the diverse activities of respondents, this re-sponse suggests that standards lack relevance for the broad rangeof tasks for which GIS is used. Since the majority of survey re-spondents work less than daily with GIS, collaborative activitiesonly partially involve GIS. The responses suggest that GIS activi-ties largely involve collaboration inside the discipline. Cross-sec-tor collaboration is limited to a few individuals whose workconsists of network maintenance and project management.

Interview MethodologyThe interviews focus on issues that provide insight into the actornetwork relationships. Issues identified in the explorative studywere the foundation for questions that explored the multifaceteddimensions of coordination. We selected three of those inter-viewed from the public, the teaching and research, and the pri-vate sectors who had indicated a willingness to participate in aninterview and who had fully completed a survey. Nine interviews,each 2 to 3 hours in duration, were held within 3 months ofmailing the initial survey.

The interviews followed semi-structured approaches; ques-tions based on the initial exploratory survey were prepared usingour evaluation of the exploratory survey. We tape-recorded andtranscribed the interviews in French. The semi-structured ap-proach provided detailed insights into perspectives of the inter-viewed person on activities and practices of actor networks.Methodologically, the most difficult aspect is assessing the role oftechnologies in actor network terms. Obviously, technologicalartifacts do not answer interview questions. In the interviews, welaid great emphasis on questions about the use and problemsassociated with using GIS technology, particularly regarding therole of standards for data sharing. From the different responses,we can interpolate the varying roles of artifacts for groups withindifferent groups.

Issues Identified in the Exploratory StudyAn analysis of the exploratory survey leads to the identificationof several issues to pursue through interviews. Determining char-acteristics of diverse relationships in actor networks includes iden-tifying degrees of participation and alliance building in networks,motivations for participating in conferences and associations, andthe role of technologies in these relationships. While the explor-ative analysis offers insights into the general characteristics ofpersons working with geographic information, some ambiguousand surprising indications of possible tensions in collaborativeactivities emerged. The low importance of standards in relationto data sharing is a case in point.

The following questions were raised in interviews:� What is your function and role in the organization where

you work?� How many people work in your organization?� What kind of work do you do and for whom?� How many people work on the same project?� What kind of relationship does your work involve with

people outside your organization?� What software do you use?� How did you choose this software?� How does the choice of software affect your work with other

organizations?� How do you share data?� What role do standards play for you now?� What role can standards play in the future?� Are technical or organizational problems more important?� How do you coordinate you work?


� What kinds of networks do you think exist between groupsusing GIS?

From these questions, we chose to focus on discussions sur-rounding cross-sector network relationships. Because of the smallnumber of interviews, we were unable to describe actual networkconnections in great detail or to better determine the types ofnetwork relationships in which people engaged. Instead of tryingto construct a model of a particular set of network relationshipsin detail, we chose to focus on issues that offered insights regard-ing the manner in which networks are constructed and main-tained.

Theses about GIS Actor NetworksThese issues guided the formulation of a set of theses re-

garding the GIS actor networks. Methodologically, the postu-lates, the hypothetical dimensions of the actor networks, thearticulated relevant aspects of the actor networks, and the con-crete issues identified by the explorative study were identified asareas to focus on in interviews. The following list provides anoverview of theses formulated before the interviews:

� Networks are relatively stable for small groups of individu-als.

� Relationships with public administration agencies dominatenetworks.

� Professionals develop GIS skills to improve specific abilitiesand projects as needed.

� Professionals have very diverse work.� Standards are not very relevant because of perceived au-

tonomy loss.� The utilization of GIS software influences relationships, and

relationships influence the utilization of technology.

Assuming that the theses adequately describe the actor net-works, an image of the GIS actor networks as an archipelagowhose islands are not all equally reachable renders a useful spatialmetaphor. Persons from islands on one end of the archipelagotend to stick together and have few exchanges with persons fromother islands. These limited actor networks promote group ac-tivities, but constrain participation to the established hierarchy.The ability to overcome hierarchies and “distance” in the actornetwork archipelago is enhanced by sharing technology with “dis-tant” individuals.

Interview ResultsThe results are insightful into the relationships that those

interviewed engaged in to coordinate a variety of GIS activities.They are also thought-provoking and indicative of the need forcontinued research. The following overview presents key charac-teristics of the actor networks. The findings are then discussed inrelation to the key issues of standards and data exchange, actornetwork alliances, and the role of technologies.

Characteristics of Actor Networks

� Actor network stabilityThere is no single network that connects all GIS profession-als, but diverse networks are constructed and disbanded asneeded. Contacts are developed in accordance with applica-tion perspectives. Individuals develop GIS in the most con-venient fashion following established users. There is atendency to scorn other individuals and groups who use adifferent GIS.

Each network is effectively stable as long as shared groupperspectives do not substantially diverge. Continual devel-opment of the networks allows change that otherwise maythreaten to destabilize the network.

� Network domination by public administrationNetworks are oriented by contacts to cantonal (state) of-fices. Contacts with federal and municipal offices are lessimportant. Academic institutions regularly conduct researchfor the cantons. Private offices work for and receive datafrom the cantons. Communities have data of interest for thecanton and are often better equipped. The federal govern-ment provides data and special skills.

� Diversity of activitiesGIS is just one area of the participants’ wide-ranging activi-ties.

� Data exchange problems are importantIn the eyes of those being interviewed, data exchange prob-lems are relative because they are so common. In practice,the respondents find a solution, yet “wish” there were globalsolutions. “Global” means standards for everybody. How-ever, none of those interviewed want to adopt standards fromoutside their discipline. External standards constrain flex-ibility, reduce independence, and allocate power, yet wouldease collaboration if others would adopt them.

� The role of technologiesThe existing use of technologies in a network has a stronginfluence on the actions of people. The multiple networksthat people belong to reinforce the use of particular arti-facts, and artifacts become distinctive characteristics of thenetwork.

The archipelago metaphor is a good simplification of actornetwork relationships, but is restrictive because the actor net-works do not take place in three-dimensional space, but in n-dimensional space, where each dimension is a relationship betweenactors. Additionally, technologies are also active and not meretools.

The interviews verified the theses, but needed to be slightlymodified:

� Networks are stable for small groups of individuals.� Relationships with public administration agencies dominate

networks.� Professionals develop GIS skills to improve specific abilities

and projects as needed.� Professionals have very diverse work.


� Standards are divisive.� The utilization of GIS software influences relationships, and

relationships influence the utilization of technology.

One thing we found unusual in the interviews and that var-ies from most descriptions of networks in terms of the stabilitywe find today is a misplaced idea that suggests static rigidity inthe face of change. Actor networks of humans and non-humansare not physically secured and maintained in the same manner aspipelines, sewers, telephone cables, and other technical elements.They are flexible, dynamic, and highly diverse. They are as un-stable as their societies. Instead of a “network” metaphor, “web”is a better metaphor for social organization defined in terms ofthe proximity between nodal points. This is not geographic prox-imity, but social proximity: an affinity between actors that corre-sponds to their interests (for various reasons) in maintaining andsupporting collective or collaborative activities with others. Websare as dynamic as the individuals, groups, artifacts, and outsidersthat change.

At the same time, the results of the interview suggest thatactor networks are not wildly oscillating. Disciplines remain ho-mogeneous and install gate-keeping procedures to retain and ex-pand their influence and income. Change is extremely slowbecause it must be digested by the entire network and acted onby a few power wielders before it affects individual practice.

The networks are correspondingly not very mobile, but aresubject to slow and certain change through the passing of time.Artifacts appear to play an important role in refiningintradisciplinary relationships, but their role in relationships in-volving broader circles would seem more divisive. Comments fromthose interviewed suggest that artifacts play a more importantrole as a significant idea in these relationships. This is an interest-ing idea, for it implies strong symbolic associations with arti-facts, but one whose examination lies outside the context of thisresearch project.

Key Aspects Identified in the SurveyThe results of research regarding key GIS aspects identified inthe explorative survey are described below.

Standards and Data ExchangeData exchange is an issue rife with conflict. This study of actornetworks examined data exchange as a key activity in terms ofstandards. In Switzerland, concerns with data exchange have beenapproached most visibly through standards (Keller and Thalmann1999). The limiting of data exchange to a matter that will beresolved on the basis of standards is a prevalent attitude amongGIS experts (Interessengemeneinschaft ARC 1997, Harvey 1998,Albrecht 1999). Although standards are sought after to improveinterdisciplinary relationships, this research suggests that stan-dards are not blanket agreements but a part of the pragmaticsolutions involving GIS in extremely varied activities. Withoutnational or international standards, the use of GIS appears tooffer advantages to many people. In fact, most of those inter-

viewed suggested that differences are part of the daily routineand are to be expected. A town surveyor stated that, “Data ex-change is at the heart of everything, but not necessarily a prob-lem.” Agencies in his town defined “rules” for exchange. Hemaintains that, in the future, these “rules” will become quite im-portant. They will become conventions. In contrast, a privateconsultant said that standards are not an issue at all. For him, themain issue is coordination; however, this has been unsuccessfulbecause agencies with resources (i.e., large utilities) work in anautonomous manner that does not consider the needs of groupswith fewer resources. These latter groups are forced to struggle tokeep up. Regarding standards, the private consultant noted a sub-stantial difference between de facto standards created in practiceand top-down “designed” standards. Designed standards fail orare out of date. De facto standards “work” and are an improve-ment. An example of a de facto standard is the DXF format. Thisformat is used by surveying engineers in federal agencies to ex-change data. The cadastre standard INTERLIS is of possible rel-evance in the future, but it is not sure if it will be widely adopted.Another viewpoint expressed was that of a private consultantworking for a GIS software company who stated that “standardsare necessary.”

The term “convention” emerged several times as the key con-cept in the creation and use of standards for data exchange. Itsuse was contingent on the disciplinary and employment of thoseinterviewed, an openness toward a federated approach, and theirassessment of the importance of forming a strategic alliance. Thoseinterviewed who sought to involve other agencies in GIS at thesame administrative level spoke overwhelmingly in a positive lightabout conventions. One surveyor described standards in this caseas mutually agreed upon “sets of rules.” Standards that were notagreed upon were perceived by those interviewed as “edicts” andas being political-economic means to enforce market positions.One consultant from a GIS software company clearly stated that“it is no longer possible to work without standards.” Others maynot dispute that but insist on using their own standards.

These results demonstrate that locally developed GIS stan-dards play an essential role in the development of actor networksby becoming critical for the enrollment of actors and enhancingpositions. One person stated during a discussion of data exchangethat standards facilitate associations. These associations are po-litical and strategic. A standard that is adopted by half of theagencies augments the importance of the group behind the stan-dard. Groups who are included in the development of standardssee how they enhance their respective actor network positions,which can lead to improved social and economic relationships.These groups become critical to arrangements betweenmultidisciplinary groups. There are technical advantages to hav-ing diverse groups agree to a convention, however, politically,these advantages accrue to the associates and members of thegroup. The proponents of GIS standards follow politics that willmake them a key player in mediating the activities of standardsand interdisciplinary collaboration. While the results of this sur-vey do not permit making any statements about inter-organiza-


tional data sharing, it provides a framework for exploring dataexchange in its political and strategic dimensions.

Alliances, Allegiances, and CollaborationAs the responses showed, GIS activities make up only a portionof the respondents’ activities. GIS is an integral part, but not thesine qua non, of work. The remarks of those interviewed suggestthat GIS is interwound in networks constructed and maintainedaccording to politically strategic social relationships. Collabora-tion takes place in a web of affiliations and long-term relation-ships between agencies that, in some cases, have existed longerthan a generation. While the interviews did not provide insightinto the historical contexts, one person interviewed who worksfor a state agency discussed the relevance of diverse representa-tions of the same object (i.e., a building) due to disciplinary dif-ferences. In his eyes, this was done because of the emphasis in thelast century on design that has influenced different disciplines invarying degrees. According to him, a big reason for this is educa-tion and training, but it is also substantially influenced by thelong-term historical roles that each discipline takes up in the ad-ministration.

In this historical institutional context, developing alliancesis a complicated activity. Even with an excess of historical hubris,actor networks are constrained although not obstructed by pastrelationships. The need to learn new technologies leads to rela-tionships between actors that deviate from prior organizationalrelationships. Private consultants play a key role in helping agen-cies move beyond their organizational confines. Young consult-ing groups are frequently upsetting older companies through theinnovative use of GIS technologies. From this comes contracts,but this leads to tensions in previously established relationships.For instance, surveyors receive criticism from those in govern-ment positions because of the surveyors’ slow adoption of GISand, therefore, a slowing down in the development of state GIS.Pornon (1992) reports that conflicts between surveyors and com-puter specialists are common in Europe.

Those interviewed describe relationships in terms of disci-plinary networks. Their comments reflect a strong division oflabor and settled disciplines in their work. Their perceptions ofalliances and, accordingly, of actor networks vary. One privateconsultant interviewed stressed the importance of the state ac-tively promoting cooperation between different public and pri-vate groups. This is contradicted in an earlier statement that thestate should play a less active role in coordinating different groups.The consultant desired a reduced state role, accompanied by anincreased emphasis on private groups fulfilling government ac-tivities. His contacts with the state underlie this interpretation.To be in the right place at the right time, it is necessary to workon building alliances that will insure his involvement in govern-mental GIS activities, regardless of whether the state privatizesGIS activities or not.

GIS technologies also lead to tensions with other groups.The development of GIS in a state in Switzerland illustrates thiswell. Several of those interviewed spoke appreciatively of the co-

ordination that the state of Vaud developed between 1985 and1995. They also pointed out that this coordination was largelyinternal. Collaboration with groups outside the state administra-tion was weak and became a contested terrain. One person inter-viewed from a state agency stated that the organized activitiesencouraged freeloading on the state. Still, he suggested that thepolitical leadership lies with the state. The state continues to oc-cupy a central position because of the number of contracts that itprovides. Its role in defining public policy is another critical rea-son. This process and the use of GIS technologies are not with-out dispute, making any project with GIS part of a broaderpolitical arrangement. The actor networks that those intervieweddescribe between individuals in governmental, state, and educa-tion sectors are only a small part of the myriad web of relation-ships connecting people and agencies.

A curious finding regarding coordination between groups isthe limited number of allegiances that are formed. Small groupsof persons only agree to cross-disciplinary allegiances that haveamply displayed their allegiance to their specific discipline; of-ten, many years in the discipline and professional position arerequired. Most of those interviewed have few contacts outsidetheir discipline. When queried, they remarked that the same ap-plied to their colleagues. The higher an individual sits in the so-cial hierarchy of the discipline, the more important they becomein arbitrating external influences, intradisciplinary relations, andin maintaining the profile and position of that particular disci-pline.

Persons who join networks have various motives and con-tribute to the network for various reasons. In the constructionand collaboration of networks, technical artifacts can play animportant role. By facilitating the definition of group member-ship, they appear to find it crucially important to determine whois included or excluded. However, the degree to which inclusionor exclusion depends on prowess with a technical artifact is un-clear and may be strongly influenced by other factors.

In these spaces, relationships follow diverse rationales. Aprivate consultant will seek relationships with cantonal officialsin the hope of being recognized by state contracts. The sameconsultant may avoid a relationship with a consultant who usesthe same software in a different town because of possible compe-tition. However, he or she will seek a relationship with a consult-ant using the same software from the same town to assure thatthe competitive situation does not result in leaving personal busi-ness interests high and dry. A strategic alliance that may be soughtin this case is really an environmental strategy for dividing lim-ited resources. Of course, the rationale for “sharing” may be lostif a stronger group is developed that is competitive enough tostand on its own.

The Roles of TechnologiesThe representation of technology as tools in GIS work, whichimplicates a “user,” complicates the study of actor networks. Fromcomments with those interviewed, the practical work of GIS al-ways involves some aspects of technology. For the most part, per-


sons differentiate between technology and organization accord-ing to their use of GIS technologies. This is influenced from se-lection to use and modification by the discipline. A good exampleof the importance of disciplines in structuring perspectives onthe role of technologies can be found in the formation of usergroups. For instance, an association was formed by users of aparticular GIS who work for competing electrical power compa-nies to ensure that data can be mutually exchanged. These groupsaugment existing disciplinary dominance by insuring the involve-ment of all groups in the technology.

Data exchange is greatly complicated due to entrenched rep-resentational differences. Because of the separation of technol-ogy from social relationship, those interviewed typify technologyas a tool that is distinct to each group. The representation of GIStechnologies as a tool is politically essential for maintaining powerand position in the actor networks. The comments discussed ear-lier regarding standards underscore the importance of technolo-gies in the actor networks of GIS. The abilities to use, control, orstandardize tools are power. Alliances and collaboration oftenmanifest themselves through use of a particular technology. Thedescription of a tool is also a description of the relationships be-tween activities and the collaboration involved.

Technological artifacts are the silent partners of humans asthey construct solutions, but are crucial to relationships. Throughthe construction of certain types of technologies, a substantialamount of effort goes into reducing friction with non-humansby congealing groups around specific mutually reinforcing con-stellations of things and people. Standards and conventions thatfacilitate particular relationships and are intended to ameliorateinterdisciplinary collaboration are extremely contentious becauseof the manner in which they constrain participation and rein-force particular disciplinary perspectives. GIS technologies arenot merely tools, they are a key component of all organizationaland disciplinary relationships.

Conclusion and OutlookThis research shows that collaboration between the various ac-tors constructing and coordinating GIS activities relies on com-plex networks between different individuals and groups that isoptimally described as a “web.” The web is a dynamic arrange-ment characterized by multiple links between actors who havepreviously defined roles and affiliations. This research presents afew essential points that can be refined in future comprehensiveprojects.

These results are relevant for broad considerations of GISpractices, including diffusion and implementation. From a meth-odological pragmatic sense, actor networks can be used to iden-tify the “contexts” of GIS use. In the case of diffusion research,this context is similar but broader than Assimakopoulos’ researchinto the Greek GIS social network. The two approaches arecomplementary at this level. It is only when research turns to theactual use of GIS in its diverse social contexts that actor networksease an intensive examination of GIS design and use from theirepistemological equity between humans and non-humans. Be-

yond research, this examination of truth and knowledge also per-mits a more participatory engagement with GIS design than sys-tems engineering approaches.

Research into the practice of GIS actor networks presents aframework for jointly considering GIS technology with those whodevelop it. Coordination and collaboration rely on establishedtechnologies, by which “established” suggests more political andsocial acceptance, not necessarily the objective technical merits.Designers have implicitly considered this in data modeling andsystem implementation. Through actor network theories, it ispossible to make these aspects explicit and an element of the de-sign process.

Some methodological ideas for future actor network studiesemerge from this work. While the results of this research showthe importance of network relationships, this study fails to indi-cate particulars of collaboration and exchange (e.g., data sharingpractices). Beyond intensive participant observation of relation-ships and the roles of technologies, extensive research on diffu-sion actor networks could be conducted to identify specific actornetwork relationships using a “snowball” questionnaire approach.By beginning with one actor and having that person pass ques-tionnaires to related collaborators, researchers could identify theactual connections and collaboration between particular actors.This would permit the identification of “traces” of alliances. Basedon the interview experiences, the role of information technolo-gies can be assessed by a more ethnographic, long-term case study,that “follows the actors” (Latour 1987). The final methodologi-cal lesson is that interviews with small focus groups appear to bean invigorating alternative to interviews in order to assess actualforms and opportunities and pitfalls of collaboration.

Collaboration between diverse groups is crucial to success-ful GIS implementation and diffusion. This research shows theimportant insights into actor network approaches that help tounderstand the broader context of specific GIS tasks and func-tions. Social coordination is essential for any organization. Un-derstanding the issues involved in using information technologiescalls for methods that consider the complex relationships betweenhumans and non-humans. Actor network theories can comple-ment GIS research in a variety of ways (Harvey and Chrisman1998).

About the Author

Dr Francis Harvey's research encompasses multiple facets of geo-graphic information science, including developing ap-proaches to organizational issues that draw on behavioralistand social constructivist work. Previous projects have exam-ined GIS implementation and design issues through a socialconstructivist approach that examined the relationships be-tween private, public, and education sector actors. Currentfunded research focuses on potentials and difficulties fordeveloping the National Spatial Data Infrastructure (NSDI)at local and regional levels. His interests also include Ger-


man-American geography, interoperability issues, andcybergeographies.

The author may be contacted at the University of Kentucky,Department of Geography, Lexington, KY [email protected].

Acknowledgments

This author wishes to thank Professor F. Golay and Profes-sor G. Berthoud for their considerable support and discussionsand Caroline Velan for her able assistance. The author gratefullyacknowledges the support of the Fonds de Rechereche UNIL-EPFL. Finally, the author would like to thank the anonymousreviewers for their many helpful comments.

References

Akrich, M., 1992, The Description of Technical Objects. In W.Bijker and J. Law (Eds.) Shaping Technology/Building Society.(Cambridge, MA: The MIT Press), 205-224.

Albrecht, J., 1999, Towards Interoperable Geo-Information Stan-dards: A Comparison of Reference Models for Geo-SpatialInformation. The Annals of Regional Science, 33, 151-169.

Assimakopoulos, D., 1997, GIS Diffusion in Greece: The De-velopment of a Greek GIS Community. In M. Craglia andH. Couclelis (Eds.) Geographic Information Research: Bridg-ing the Atlantic. (London: Taylor & Francis), 111-128.

Calkins, H.W. and N.J. Obermeyer, 1991, Taxonomy for Sur-veying the Use and Value of Geographical Information. In-ternational Journal of Geographical Information Systems, 5,341-351.

Callon, M., 1986, Some Elements of a Sociology of Translation.Domestication of the Scallops and the Fishermen of St.Brieux Bay. In J. Law (Ed.) Power, Action and Belief. A NewSociology of Knowledge. (London: Routledge and Paul Kegan),196-229.

Callon, M., 1987, Society in the Making: The Study of Technol-ogy as a Tool for Sociological Analysis. In W.E. Bijker, T.P.Hughes, and Trevor J. Finch (Ed.) The Social Construction ofTechnological Systems: New Directions in the Sociology andHistory of Technology. (Cambridge, MA: The MIT Press),83-103.

Callon, M., 1991, Techno-Economic Networks and Irreversibil-ity. In J. Law (Ed.) A Sociology of Monsters: Essays on Power,Technology and Domination. (London: Routledge), 132-161.

Callon, M., J. Law, and A. Rip, 1986, Qualitative Scientometrics.In M. Callon, J. Law, and J.A. Rip (Eds.) Mapping the Dy-namics of Science and Technology. Sociology of Science in theReal World. (London: Macmillan).

Campbell, H. and I. Masser, 1995, GIS and Organizations. HowEffective are GIS in Practice? (London: Taylor & Francis).

Demeritt, D., 1996, Social Theory and the Reconstruction ofScience and Geography. Transactions of the Institute of Brit-ish Geographers, 21, 484-503.

Giddens, A., 1979, Central Problems in Social Theory: Action, Struc-ture and Contradiction in Social Analysis. (London:Macmillan).

Harvey, F., 1998, Improving Access to Geographic InformationQuality. In M. Craglia (Ed.) Proceedings of the First AGILEMeeting.

Harvey, F. and N.R. Chrisman, 1998, Boundary Objects and theSocial Construction of GIS Technology. Environment andPlanning A, 30, 1683-1694.

Interessengemeneinschaft ARC, 1997, Empfehlungen zurStandardisierung von föderalistisch verwalteten Gebäudedaten.(Bern: SIK-GIS).

Keller, S. and H. Thalmann, 1999, Modeling and Sharing GraphicPresentations of Geospatial Data. In A. Vckovski, K.E.Brassel, and H.-J. Schek (Eds.) Interoperating Geographic In-formation Systems. (Berlin: Springer-Verlag), 151-162.

Latour, B., 1987, Science in Action. How to Follow Scientists andEngineers through Society. (Cambridge, MA: Harvard Uni-versity Press).

Latour, B., 1992, Where are the Missing Masses? The Sociologyof a Few Mundane Artifacts. In W. Bijker and J. Law (Eds.)Shaping Technology/Building Society. (Cambridge, MA: TheMIT Press), 265.

Latour, B., 1993, We Have Never Been Modern. C. Porter, Trans-lator. (Cambridge, MA: Harvard University Press).

Latour, B., 1999, Give Me a Laboratory and I Will Raise theWorld. In M. Biagioli (Ed.) The Science Studies Reader. (NewYork: Routledge), 258-275.

Latour, B. and S. Woolgar, 1979, Laboratory Life: The Social Con-struction of Scientific Facts. (Beverly Hills, CA: Sage Publica-tions).

Mol, A. and J. Law, 1994, Regions, Networks and Fluids: Ane-mia and Social Topology. Social Studies of Science, 24, 641-671.

Nedovic-Budic, Z. and D.R. Godschalk, 1994, Implementationand Management Effectiveness in Adoption of GIS Tech-nology in Local Governments. Computers, Environment, andUrban Systems, 18, 285-304.

Onsrud, H.J. and G. Rushton (Eds.), 1995, Sharing GeographicInformation. (New Brunswick, NJ: Center for Urban PolicyResearch).

Pinto, J.K. and B. Azad, 1994, The Role of Organizational Poli-tics in Implementation. URISA Journal, 6, 35-61.

Pornon, H., 1992, SIG: Mise en Oeuvre et Applications. (Paris:Editions Hermes).

Rogers, E.M., 1983, Diffusion of Innovations. (New York: TheFree Press).

Schuler, D. and A. Namioka (Eds.), 1993, Participatory Design.Principles and Practices. (Hillsdale, NJ: Lawrence ErlbaumAssociates).

Mark Your Calendar!○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

w w w . u r i s a . o r g

New Specialty Conference!

Technology — MakingPublic Works

WORK BETTERMay 6-8, 2001Holiday Inn O'Hare International,Rosemont, IL

New Specialty Conference!

Caribbean GISSeptember 9-12, 2001Wyndham Rose HallMontego Bay, Jamaica

Street Smart & AddressSavvy ConferenceAugust 12-14, 2001Milwaukee, WI

URISA 2001 AnnualConference& Exposition

October 20-24, 2001Long Beach Convention CenterLong Beach, CA

URISA Journal � Literature Review 39

Review of CurrentJournal Literature

Editorial Intent

For the review of Current Journal Literature, we have selected 21 journals of relatedinterest to URISA members (see next page for list of journals). The Journal Literatureeditors scan these journals for articles they feel are relevant to the URISA audience.Selected articles are then assigned one of nine categories, which are modeled after theindex categories used for the Journal and the URISA Proceedings. Several additionalcategories have been added. The nine categories are:

� Analysis, Modeling and Simulation� Applications� Cartography� Data� Hardware, Software and Technology� Implementation and Management� Remote Sensing and GPS� System Concepts and Theory� Other Issues and Topics

Please note that there are no clear-cut boundaries between categories, and some articlesmay qualify for entry in more than one. However, we decided against repetitive entriesfor a single article, so the reader is advised to look in more than one category for aparticular entry. Also, note that some entries are list only the beginning page number ofan article. These articles were retrieved online, and ending pages were not provided.

If you have suggestions or comments about our procedures or about this section, pleasecontact us.

Zorica Nedovic´-Budic´

Special thanks go out to ZoricaNedovic´-Budic´ who again, took thetime to compile and edit this Litera-ture Review. Due to her tireless effortsand dedication to URISA we are ableto include this section as an annualfeature of the URISA Journal.

Zorica Nedovic´–Budic´ is an as-sistant professor of urban planning andgeographic information systems (GIS).She earned a Ph.D. degree from theUniversity of North Carolina atChapel Hill in 1993. Her main areaof research is the implementation ofGIS in local government settings andthe evaluation of GIS impact on ur-ban planning functions and process.Comparative study of urban develop-ment and planning practice in formercommunist countries in Europe is an-other current research interest.


Selected JournalsIJGIS International Journal of Geographical Information Science (monthly)

Taylor and Francis Ltd., 1. Gunpowder Square, London EC4A 3DE, UKSLIS Surveying and Land Information Systems (quarterly)

American Congress on Surveying and Mapping, 5410 Grosvenor Lane,Suite100, Bethesda, MD 20814-2122

CGIS Cartography and Geographic Information Systems/Science (quarterly)American Congress on Surveying and Mapping, 5410 Grosvenor Lane,Suite 100, Bethesda, MD 20814-2122

CA Cartographica (quarterly)University of Toronto Press, 5201 Dufferin Street, Downsview, OntarioM3H 5T8, Canada

PERS Photogrammetric Engineering and Remote Sensing (monthly)American Society of Photogrammetry and Remote Sensing, 5410 Grosvenor Lane, Suite 210, Bethesda, MD 20814-2160

IJRS International Journal of Remote Sensing (monthly)Taylor and Francis Ltd., 1. Gunpowder Square, London EC4A 3DE, UK

CG Computers and Geosciences (monthly)Pergamon Press Ltd., Linacre House, Jordan Hill, Oxford OX2 8DP, UK

MISQ Management Information Systems Quarterly (quarterly)Carlson School of Management, University of Minnesota, 271 19th Ave.South, Minneapolis, MN 55455

JAPA Journal of the American Planning Association (quarterly)American Planning Association, 122 S. Michigan Ave., Suite 1600, Chicago, IL 60603

EPB Environment and Planning B (bimonthly)Pion Limited, 207 Brondesbury Park, London NW2 5JN, UK

CEUS Computers, Environment and Urban Systems (bimonthly)Pergamon Press, Inc., Fairview Park, Elmsford, New York 10523

LUP Landscape and Urban Planning (bimonthly)Elsevier Science B.V., Journal Department, P.O. Box 211, 1000 A.E. Amsterdam, The Netherlands

EM Environmental Management (bimonthly)Springer-Verlag New York Inc., 175, Fifth Avenue, New York, NY 10010

JEM Journal of Environmental Management (monthly)Academic Press Ltd., 6277 SeaHarbor Drive, Orlando FL 32887-4900

PAR Public Administration Review (bimonthly)American Society for Public Administration (ASPA), 1120 G Street NW,Suite 700, Washington, DC 20005-3885

JUA Journal of Urban Affairs (quarterly)JAI Press, Inc., 55 Old Post Road No.2, Box 1678, Greenwich, CT 06836-1678

TRR Transportation Research RecordTransportation Research Board, National Research Council,2101 Constitution Avenue, Washington, DC 20418

JRS Journal of Regional Science (quarterly)Blackwell Publishers108 Cowley Road, Oxford OX4 1JF, UK350 Main Street, Malden, MA 02148, USA

JUT Journal of Urban Technology (quarterly)Carfax Publishing, Taylor & Fancis Ltd.Rankine Road, Basingstroke, Hants RG24 8PR, UK325 Chestnut Street, 8th Floor, Philadelphia, PA 19106, USA


TGIS Transactions in GISBlackwell Publishers Ltd., Blackwell Publishers Inc., Marston Book Services, PO Box 269, Abingdon, Oxon OX144YN, UK

JASIS Journal of the American Society for Information ScienceJohn Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158.

1. Analysis, Modeling and Simulation

Anton, J.J., Rompaey, V., Govers, G., and Baudet, M. 1999. A strategy for controlling error of distributed environmental modelsby aggregation. IJGIS 13(6): 577-590

Bruin, S. D. 2000. Querying probabilistic land cover data using fuzzy set theory. IJGIS 14(4): 359-372Etzelmüller, B. 2000. On the Quantification of Surface Changes using Grid-based Digital Elevation Models (DEMs). TGIS 4(2):

129-143Finco, M. V., and Hepner, G. F. 1999. Investigating US-Mexico border community vulnerability to industrial hazards: A simula-

tion study in Ambos Nogales. CGIS 26(4): 243-252Li, X., and Yeh, A. G. 2000. Modelling sustainable urban development by the integration of constrained cellular automata and

GIS. IJGIS 14(2): 131-152Liang, C., and MaCkay, D. S. 2000. A general model of watershed extraction and representation using globally optimal flow paths

and up-slope contributing areas. IJGIS 14(4):337-358Paniconi, C., Kleinfeldt, S., Deckmyn, J., and Giacomelli, A. 1999. Integrating GIS and data visualization tools for distributed

hydrologic modeling. TGIS 3(2):97-118Shi, W., and Pang, M. Y. C. 2000. Development of Voronoi-based cellular automata - an integrated dynamic model for Geographi-

cal Information Systems. IJGIS 14(5): 455-474Taylor, K., Walker, G., Abel, D., Arctur, D., and Hair, D. 1999. A framework for model integration in spatial decision support

systems. IJGIS 13(6): 553-555Wang, X., White-Hull, C., Dyer, S., and Yang, Y. 2000. GIS-ROUT: a river model for watershed planning. EPB 27(2): 231-246Yang, X., and Hodler, T. 2000. Visual and statistical comparisons of surface modeling techniques for point-based environmental

data. CGIS 27(2): 165-175

2. Applications

Bishop, I. D., and Gimblett, H. R. 2000. Management of recreational areas: GIS, autonomous agents, and virtual reality. EPB27(3): 423-435

Chakraborty, J., Schweitzer, L. A., and Forkenbrock, D. J. 1999. Using GIS to Assess the Environmental Justice Consequences ofTransportation System Changes. TGIS 3(3): 239-258

Collishonn, W., Pilar, J. V. 2000. A direction dependent least-cost-path algorithm for roads and canals. IJGIS 14(4):397-406Holtier, S., Steadman, J. P., and Smith, M. G. 2000. Three-dimensional representation of urban built form in a GIS. EPB 27(1):

51-72Jiang, H., and Eastman, R. 2000. Application of fuzzy measures in multi-criteria evaluation in GIS. IJGIS 14(2): 173-184O’Sullivan, D., Morrison, A., and Shearer, J. 2000. Using desktop GIS for the investigation of accessibility by public transport: an

isochrone approach. IJGIS 14(1): 85-104Talen, E. 2000. Bottom-up GIS: A new tool for individual and group expression in participatory planning. JAPA 66(3): 279-294Tarabanis, K., and Tsionis, I. 1999. Using network analysis for emergency planning in case of an earthquake. TGIS 3(2): 187-197

3. Cartography

Bacino, C. C. 1999. Automating the parcel mapping process: The Montana cadastral project. SLIS 59(3): 165-168Bunch, R., and Lloyd, R. 2000. The search for boundaries on maps: Color processing and map pattern effects. CGIS 27(1): 15-29Chirie, F. 2000. Automated name placement with high cartographic quality: City street maps. CGIS 27(2): 101-110Forrest, D. 1999. Geographic Information: Its nature, classification, and cartographic representation. CA 36(2): 31-53Hojholt, P. 2000. Solving space conflicts in map generalization: Using a finite element method. CGIS 27(1): 65-73


Kent, R. B., and Klosterman, R. E. 2000. GIS and mapping: Pitfalls for planners. JAPA 66(2): 189-198Sadahiro, Y. 2000. Perception of spatial dispersion in point distributions. CGIS 27(1): 51-64Tao, C. V. 1999. Error modeling and analysis of terrestrial stereo imaging systems for mobile mapping. SLIS 59(3): 187-196Zimmer, R. 1999. Creating a county mapping control network: A work in progress. SLIS 59(3): 159-163

4. Data

Ahlqvist, O., Keukelaar, J., and Oukbir, K. 2000. Rough classification and accuracy assessment. IJGIS 14(5): 475-496Arbia, G., Griffith, D., and Haining, R. 1999. Error propagation modeling in raster GIS: Adding and ratioing operations. CGIS

26(4): 297-315Argeseanu, V. S. 1999. Accuracy evaluation of the geoid height model GEOID96 for the Rocky Mountains States. SLIS 59(4):

241-256Baumgartner, A., Steger, C., and Mayer, H. 1999. Automatic road extraction based on multi-scale, grouping, and context. PERS

65(7): 777-785Carson-Lambert, S., and Garie, H. 1999. The National Spatial Data Council: a true partnership for the NSDI. PERS 65(11):

1231-1238Fuller, G. W. 1999. A vision for a Global Geospatial Information Network (GGIN) creating, maintaining and using globally

distributed geographic data, information, knowledge and services. PERS 65(5): 524-525Gelbman, E., and Doytsher, Y. 1999. Automatic filtering and classification of raw field surveying data. SLIS 59(4): 231-240Harvey, F. J., Buttenfield, B. P., and Carson-Lambert, S. 1999. Integrating geodata infrastructures from the ground up. PERS

65(11):1287-1291Kelly, N. M. 2000. Spatial accuracy assessment of wetland permit data. CGIS 27(2): 117-127Kyriakidis, P. C., Shortridge, A. M., and Goodchild, M.F. 1999. Geostatistics for conflation and accuracy assessment of digital

elevation models. IJGIS 13(7): 677-707Larson, K., Burton, G., Scarrah, P., and Snyder, B. 1999 Don’t duck metadata. SLIS 59(3): 169-173Leitner, M., and Buttenfield, B. P. 2000. Guidelines for the display of attribute certainty. CGIS 27(1): 3-14Longley, P. A., and Harris, R. J. 1999. Towards a new digital data infrastructure for urban analysis and modeling. EPB 26(6): 855-

878López, C. 2000. Improving the Elevation Accuracy of Digital Elevation Models: A Comparison of Some Error Detection Proce-

dures. TGIS 4(1): 43-64Mates, D. T. 1999. The BLM’s GCDB: A shared database. SLIS 59(3): 155-157McKee, L. 1999. Breakthrough in Web-based geographic information. PERS 65(11): 1239-1285Shi, W., and Liu, W. 2000. A stochastic process-based model for the positional error of line segments in GIS. IJGIS 14(1): 51-66Theobald, D. M. 2000. Reducing linear and perimeter measurement errors in raster-based data. CGIS 27(2): 111-116Tudor, G. S., and Wolfe, C. 1999. Washington State cadastral framework project: Implementing the FGDC cadastral data content

standard and integrating data from multiple sources. SLIS 59(3): 179-185Wang, K., and Lo, C. 1999. An assessment of the accuracy of triangulated irregular networks (TINs) and lattices in ARC/INFO.

TGIS 3(2): 161-174Wise, S. 1999. Extracting raster GIS data from scanned thematic maps. TGIS 3(3): 221-237

5. Hardware, Software and Technology

Brown, I. 1999. Developing a virtual reality user interface (VRUI) for geographic information retrieval on the Internet. TGIS3(3):207-220

Churcher, C., and Churcher, N. 1999. Realtime conferencing in GIS. TGIS 3(1): 23-30Evans, J. D. 1999. Interoperable Web-based services for digital orthophoto imagery. PERS 65(5): 567-571Li, B. 2000. A component perspective on geographic information services. CGIS 27(1): 75-86


6. Implementation and Management

Assimakopoulos, D. G. 2000. Social network analysis as a tool for understanding the diffusion of GIS innovations: the Greek GIScommunity. EPB 27(4): 627-640

Bharadwaj, A. S. 2000. A resource-based perspective on information technology capability and firm performance: An empiricalinvestigation. MISQ 24(1): 169-196

Brown, C. V. 1999. Horizontal mechanisms under differing IS organization contexts. MISQ 23(3): 421-54Chan, T. O., and Williamson, I. P. 2000. Long term management of a corporate GIS. IJGIS 14(3): 283-303Cooper, R. B. 2000. Information technology development creativity: A case study of attempted radical change. MISQ 24(2): 245-

276Craglia, M., and Signoretta, P. 2000. From global to local: the development of local geographic information strategies in the

United Kingdom. EPB 27(5): 759-775Gregor, S., and Benbasat, I. 1999. Explanations from intelligent systems: Theoretical foundations and implications for practice.

MISQ 23(4): 497-530Hendricks, P. H. J. 2000. An organizational learning perspective on GIS. IJGIS 14(4): 373-396Ikhuoria, I. A. 1999. Professionalism and manpower issues in geographical information systems in West Africa. TGIS 3(4): 343-

358Sheppard, E., Couclelis, H., Graham, S., Harrington, J. W., and Onsrud, H. 1999. Geographies of the I nformation society. IJGIS

13(7): 797-823.Man, W. H. 2000. Institutionalization of Geographic Information Technologies: Unifying concept? CGIS 27(2): 139-151Montagu, S. 2000. GIS and natural resource planning in Papua New Guinea: a contextual analysis. EPB 27(2): 183-196Montealegro, R., and Keil, M. 2000. De-escalating information technology projects: Lessons from the Denver International

Airport. MISQ 24(3): 417-447Moore, J. E. 2000. One road to turnover: A examination of work exhaustion in technology professionals. MISQ 24(1): 141-168Nambisan, S., Agarwal, R., and Tanniru, M. 1999. Organizational mechanisms for enhancing user innovation in information

technology. MISQ 23(3): 365-395Nedovic-Budic, Z., and Pinto, J. F. 2000. Information sharing in an interorganizational GIS environment. EPB 27(3): 455-474Oliver, S. G. 1999. The role of states as key stakeholders in the National Spatial Data Infrastructure: Where the rubber meets the

road. PERS 65(11): 1301-1302Ramasubramanian, L 1999. GIS implementation in developing countries: Learning from organisational theory and reflective

practice. TGIS 3(4): 359-380Ravichandran, T., and Rai, A. 2000. Quality management in systems development: An organizational system perspective. MISQ

24(3): 381-415Reich, B. H., Brown, K., and Lynn, M. 1999. “Seeding the line”: Understanding the transition from IT to non-IT careers. MISQ

23(3): 337-364Reich, B. H., and Benbasat, I. 2000. Factors that influence the social dimension of alignment between business and information

technology objectives. MISQ 24(1): 81-113Swanson, E. B., and Dans, E. 2000. System life expectancy and the maintenance effort: Exploring their equilibration. MISQ 24(2):

277-297Wastell, D. G. 1999. Learning dysfunctions in information systems development: Overcoming the social defenses with transi-

tional objects. MISQ 23(4): 581-600Weill, P., and Vitale, M. 1999. Assessing the health of an information systems applications portfolio: An example from process

manufacturing. MISQ 23(4): 601-624


7. Remote Sensing and GPS

Cohen, G. 1999. Differential global positioning systems: Determination of Franklin-Lebanon town line, fall 1998. SLIS 59(3):203-204

Heo, J., FitzHugh, T. W. 2000. A standardized radiometric normalization method for change detection using remotely sensedimagery. PERS 66(2): 173-181

Jensen, J. R., and Cowen, D. C. 1999. Remote sensing of urban/suburban infrastructure and socio-economic attributes. PERS65(5): 611-622

Kelly, M., Estes, J. E., and Knight, K. A. 1999. Image interpretation keys for validation of global land-cover data sets. PERS 65(9):1041-1050

McCormick, C. M. 1999. Mapping exotic vegetation in the Everglades from large-scale aerial photographs. PERS 65(2): 179-184Mikhail, E. M. 1999. Is photogrammetry still relevant. PERS 65(7): 740-741Parks, W. 1999. Accuracy of GPS-derived orthometric height in San Diego County, California. SLIS 59(4): 257-275Shi, W. Z., Ehlers, M., and Tempfli, K. 1999. Analytical Modelling of Positional and Thematic Uncertainties in the Integration of

Remote Sensing and Geographical Information Systems. TGIS 3(2): 119-136

8. System Concepts and Theory

Albert, W. S., and Golledge, R. G. 1999. The use of spatial cognitive abilities in geographical information systems: The mapoverlay operation. TGIS 3(1): 7-21

Bibby, P., and Shepherd, J. 2000. GIS, land use, and representation. EPB 27(4): 583-598Chakroun, H., Benie, G. B., O’Neill, N. T., and Desilets, J. 2000. Spatial analysis weighting algorithm using Voronoi diagrams.

IJGIS 14(4):319-336Cheng, Y., and Lorre, J. J. 2000. Equal area map projection for irregularly shaped objects. CGIS 27(2): 91-100Chrisman, N. R. 1999. What does ‘GIS’’ mean? TGIS 3(2): 175-186Dragicevic, S., and Marceau, D. J. 2000. A fuzzy set approach for modeling time in GIS. IJGIS 14(3): 225-245Egenhofer, M. J., Glasgow, J., Gunther, O., Herring, J. R., and Peuquet, D. J. 1999. Progress in computational methods for

representing geographical concepts. IJGIS 13(8): 775-796Gunther, O., Picouet, P., Sanglio, J.M., Scholl, M., and Oria, V. 1999. Benchmarking spatial joins a la carte. IJGIS 13(7): 639-655Hornsby, K., and Egenhofer, M. J. 2000. Identity-based change: a foundation for spatio-temporal knowledge representation. IJGIS

14(3): 207-224Malczewski, J. 2000. On the Use of Weighted Linear Combination Method in GIS: Common and Best Practice Approaches. TGIS

4(1): 5-22Mark, D. M., Freksa, C., Hirtle, S. C., Lloyd, R., and Tversky, B. 1999. Cognitive models of geographical space. IJGIS 13(8): 747-

774Mrozinski, R. D., Cromley, R. G. 1999. Singly- and doubly-constrained methods of areal interpolation for vector-based GIS. TGIS

3(3): 285-301Nackaerts, K., Govers, G., and Orshoven, J. V. 1999. Accuracy assessment of probabilistic visibilities. IJGIS 13(7): 709-721Nicholas, C. 1999. A transformational approach to GIS operation. IJGIS 13(7): 617-637.Saalfeld, A. 1999. Delaunay triangulations and stereographic projections. CGIS 26(4): 289-296Stefanakis, E., Vazirgiannis, M., and Sellis, T.1999. Incorporating fuzzy set methodologies in a DBMS repository for the applica-

tion domain of GIS. IJGIS 13(7): 657-675Veregin, H. 1999. Line simplification, geometric distortion, and positional error. CA 36(1): 25-39Veregin, H. 2000. Quantifying positional error induced by line simplification. IJGIS 14(2): 113-130Wilcox, D. J., Harwell, M. C., and Orth, R. J. 2000. Modeling dynamic polygon objects in space and time: A new graph-based

technique. CGIS 27(2): 153-164Winter, S. 2000. Uncertain topological relations between imprecise regions. IJGIS 14(5): 411-430Woodcock, C. E., and Gopal, S. 2000. Fuzzy set theory and thematic maps: accuracy assessment and area estimation. IJGIS 14(2):

153-172Yuan, M. 1999. Use of a three-domain repesentation to enhance GIS support for complex spatiotemporal queries. TGIS 3(2): 137-

159


9. Other Issues and Topics

Goodchild, M. F., Egenhofer, M. J., Kemp, K. K., Mark, D. M., and Sheppard, E. 1999. Introduction to the Varenius Project. IJGIS13(8): 731-745

Greenfeld, J. S. 2000. Surveyors and GIS – The professional and educational challenges. IJGIS 60(1): 7-12

Think about your coworkers and peers.

Think about your IT vendors.

Think about students who aspire to work

in the field. And tell them about URISA!

Did you know that most people join

URISA because a member like you en-

couraged them to join?

Last year, URISA introduced thisspecial series of publications

designed to help professionals in needof important information, but withoutthe time to read extensive literature orattend expensive seminars. Because oftheir overwhelming success, URISA ispleased to introduce three new titles.

All of these tutorial-style books providethe reader with…� Subject matter fundamentals� Term definitions� Negotiating and purchasing tips� Recommendations� Flowcharts, diagrams, photos and

other examples

All Quick Study publications are from20-40 pages in length and are preparedby recognized experts in their fields.

GIS Glossary ofTermsEdited by Daniel Parr

This handy reference guide was written inorder to equip the IT/GIS professionalwith the definitions, terms and conceptsthat are central to the GIS community. Inorder to have a successful GIS, onemust be familiar with the terms andsyntax that are its basis. The GISGlossary contains hundreds ofdefinitions that pertain to GIS, IT, andrelated fields. The terms and conceptsdefined will help to provide a solidknowledge base to the novice GIS user,as well as to bolster the working

vocabulary of even themost seasoned GISprofessional.#511, Member $15,Nonmember $19

Internet BasedGeographicInformation Systemsand DecisionSupport ToolsBy Shilpam Pandy

Rapid growth of the Internet over the pastdecade has opened up exciting new waysto supply data, tools, models and otherinformation to potential users. Internetdelivery provides opportunities toincrease involvement of stakeholders inthe decision-making and planning processby providing knowledge and data througha widely accessible, fast, cost-effective,and easy-to-use medium.

This Quick Study provides readers with anunderstanding of the availability andpotential of existing Internet resources,and prepares them to use the nextgeneration of Internet based decisionsupport tools. Using specific and realworld examples this important publicationdiscusses the approaches and challengesinvolved in setting up and implementing adecision support tool application. Thepublication includes sections on:

� Advantages of Internet based tools� Serving web-based GIS and decision

support tools� Software development platform

examples� Sample implementation of an Internet

based decision support tool� Mapping and user interface issues� Internet GIS terminology

This new book is both for professionalsinterested in learning about the types ofInternet based decision support toolscurrently available, and for those

individuals involved in developing orextending geographic informationand analytical capabilities over theInternet.#555, Member $15,Nonmember $19

Quick Study Series“When you need the information, butdon’t have the time!”

The Internet,Web, andE-Commerce:A guide to assisting small andmid-size organizations inproviding on-line servicesBy David Martin and David Reid

This is URISA’s first publication aimedspecifically at the needs of small and mid-sized organizations. This Quick Study willprovide you with the informationnecessary to 1) get your organization on-line, 2) develop a useful World Wide Website and 3) begin conducting commercevia electronic transactions (e-commerce).This publication is extremely valuable as aresource in designing a simple, yetcomprehensive plan to execute yourorganization’s on-line strategy.

� Learn the options available to youregarding networking and Internetconnections.

� Find out the 20 questions you needto ask your Internet service providerto ensure you are getting the bestservice for your money.

� Learn the do’s and don’ts of websitedevelopment including the 10 designpitfalls you must avoid.

� Find out the key factors that anorganization should consider whendevising an e-government/businessstrategy.

This timely new publication also containsrelevant case studies, term definitions,partnership options, staffing issues,software and hardware options and linksto dozens of websites for furtherinformation and examples. Whether youjust need to know enough to makeinformed decisions or you are chargedwith developing a complete Internet/Web/E-Commerce solution yourself, you’ll findthis book your most referred to source ofinformation on this important topic.#554, Member $15, Nonmember $19

Visit our complete publications libraryonline at www.urisa.org/store

URISA Journal � Technology Reports 47

Information TechnologyReports and Studies: Update

From time to time the URISA Journal intends to introduce and recommend reportsand studies that may be of interest to significant portions of the URISA membershipyet may have had limited exposure to date. Such reports will be drawn from the pool ofstudies and reports issued each year by government agencies at all levels, workshopreports, and studies accomplished by related professional organizations. At times wewill reprint only an executive summary while at other times we may reproduce theentire text of an important report.

Introductions to recent reports and studies of relevance to URISA members in theremainder of this issue include the following:

1. Know the Rules. Use the ToolsPrivacy in the Digital Age: A Resource for Internet UsersPrepared by U.S. Senate Committee on the Judiciary

2. NSDI Community Demonstration Projects: Final ReportPrepared by the Community, Federal and Business Partners of the NSDI Commu-nity Demonstration Projects

3. 2000 E-Government SurveyPrepared by National Association of Counties and Public Technology, Inc.

4. Data for Science and Society: The Second National Conference on Scientific andTechnical DataU.S. National Committee for CODATA, National Research Council

5. Declaring Independence: A Guide to Creating Community-Controlled ScienceJournalsThe Scholarly Publishing and Academic Resources Coalition (SPARC)

The following 5 reports were com-piled by the Editor-in-Chief of theURISA Journal Harlan J. Onsrud.Dr. Onsrud is Professor of SpatialInformation Science and Engineer-ing at the University of Maine anda research scientist with the Na-tional Center for Geographic Infor-mation and Analysis (NCGIA). Heteaches courses in information sys-tems law, cadastral and land infor-mation systems, environmental lawand land development design. Dr.Onsrud's research focuses on theanalysis of legal, ethical, and insti-tutional issues affecting the creationand use of digital spatial databasesand the assessment of the socialimpacts of spatial technologies.


Know the Rules, Use the ToolsPrivacy in the Digital Age: A Re-source for Internet UsersPrepared by U.S. Senate Committee on the Judiciary

Editors Note: Following is the Table of Contents and ExecutiveSummary for this Report. The entire report may be found on-line at http://judiciary.senate.gov/privacy.htm

TABLE OF CONTENTS

LETTER FROM SENATOR ORRIN G. HATCH, CHAIR-MAN,U.S. SENATE COMMITTEE ON THE JUDICIARY

EXECUTIVE SUMMARY

I. INTRODUCTION

II. THE ONLINE PRIVACY ISSUE

A. What is On-line PrivacyB. An Old Problem With a New TwistC. Online Privacy Distinguished From Online SecurityD. What are “Cookies” and How Do They Impact Online

PrivacyE. What Do Consumers Think About Online Privacy

III. WHAT IS BEING DONE TO ADDRESS CONCERNSABOUT ONLINE PRIVACY?

A. Senate Judiciary Committee WorkB. What Progress is Being Made by Industry in Protecting

Online PrivacyC. The Need to Empower Consumers to Protect Their

PrivacyD. What Can Consumers Do to Protect Their Privacy

IV. RESOURCES

A. Technologies Available to Consumers1. Ways of Handling Cookies

a. Internet Browser Settingsb. Manual Deletion of Cookies Using Browser

Filesc. Cookie-Cutters

2. Identity Scrubbersa. Privada Controlb. Incogno SafeZonec. Freedomd. Anonymizer.come. Crowds

1. Privacy Preference Technologya. AT&T Researchb. Privacy Right

2. Digital Identity Managers/Preference Organizersa. Microsoft Passportb. Iprivacy Identity Managerc. Digitalme

3. Infomediariesa. Orby Privacy Plusb. Persona Inc.c. Lumeriad. Respond.com

4. Permission Marketing

5. Business to Business Technologiesa. Ad Delivery Servicesb. Customer Relationship Management

6. Impermanent Email Technologya. Disappearing Effect

B. Website Seal Programs1. Trust e.2. BBBOnline3. CPA WebTrust

C. Organizations Involved in Online Privacy Dialogue1. Center for Democracy and Technology2. Direct Marketing Association3. Electronic Privacy Information Center4. Internet Alliance5. Online Privacy Alliance6. Platform for Privacy Preferences7. Privacy Rights Clearinghouse

V. CONCLUSION

The Senate Judiciary committee, chaired by Senator Orrin G. Hatch,intends to continue to examine and develop public policy that con-siders the needs of consumers, law enforcement and online businesses;ensures continued investment in technological development in thisimportant area; and enables e-commerce to reach its full potential.In furtherance of this goal, and in an effort to provide timely infor-mation, Senator Hatch has issued the following resource guide forconsumers.

With online sales in the billions of dollars, it is clear thatmany consumers have embraced the digital marketplace. Retailelectronic commerce (or “e-commerce”) sales reached an astound-ing $5.3 billion in the first quarter of 2000. Analysts predict that


online shopping could grow to $78 billion per year by 2003.With more than a third of all households in the United Statesonline, consumers increasingly are using the Internet to conductdaily activities ranging from filling their prescriptions to personalbanking to trading stocks. At the same time, businesses have rec-ognized the tremendous potential the digital marketplace offers.

Yet, as Microsoft, a leader of the new economy, stated in anadvertisement entitled “Happy e-Holidays,” it is apparent that“not everyone is sold on Internet shopping. Still, many peopleare worried about sharing their credit card and other personalinformation over the Internet.” The Senate Judiciary Committeestaff finds that certain conduct is taking place online that couldthreaten to chill the continued rapid expansion of the digitalmarketplace: the extensive collection by websites of personallyidentifiable information about consumers, often without con-sumers’ consent or knowledge.

� First, Committee staff finds that many consumers are un-aware that personally identifiable information is being col-lected about them while they surf the Net. For example, arecent study found that among heavy Internet users, 12percent were uncertain about what a “cookie” was (a cookieis an electronic tag placed on an individual’s hard drive byan Internet site to identify the individual while he surfsthe Internet).

� Second, Committee staff finds that consumers are concernedabout the collection and use of personally identifiable infor-mation. A study found that 87 percent of individuals usingthe Internet are concerned about threats to their personalprivacy.

� Third, Committee staff finds that most consumers are notaware of technological tools and resources that are availableto empower them to protect their privacy.

� Fourth, Committee staff finds that the expansion of e-com-merce may be jeopardized if consumer concerns are not ad-equately addressed. A study conducted by the NationalLeague of Cities found that among Internet users who re-search products or services online (42 percent of all Internetusers), only 24 percent actually purchase products or ser-vices online. The same study found that 73 percent of Inter-net users are not comfortable providing credit card orfinancial information to businesses online, and 70 percentare not comfortable providing personal information.

Most Websites Collect Users’ Information

The vast majority of websites collect personally identifiableinformation from consumers. Increasingly, these websites areposting privacy policies —statements that inform the consumerof the type of information the website collects and to whom suchinformation is sold. Moreover, a number of individual compa-nies have taken steps to respond to privacy concerns by askingfor affirmative consent from consumers before collecting and sell-ing personally identifiable information.

Consumer Education is Critical

To protect individual privacy online, consumers must un-derstand at the outset whether websites collect personally identi-fiable information, and if so, the extent to which the websites usepersonally identifiable information. Consumers also must un-derstand whether websites implement privacy policies, and if so,whether those policies protect or compromise individual privacy.In addition, consumers need an awareness of the various resourcesand technological tools that are available to them for protectingtheir online privacy. With this knowledge, consumers can makean informed judgment about whether to provide informationrequested by a particular website.

Resources Available to Consumers

A number of resources are available to consumers who wantto protect their online privacy. Groups involved in the debateover online privacy, such as the Direct Marketing Association,the Electronic Privacy Information Center, the Internet Alliance,the Online Privacy Alliance, the Platform for Privacy Preferences,and the Privacy Rights Clearinghouse, provide helpful informa-tion. Consumers may further protect their privacy online by uti-lizing websites that adhere to “seal programs,” such as TRUSTe,BBBOnline, CPA WebTrust, and Enonymous.com, which areindependent, third-party organizations that monitor and/or ratethe privacy practices of websites.

Technology Tools Can Empower Consumers

Consumers can be empowered to protect their privacy byunderstanding the technological tools that are available to safe-guard their personally identifiable information. For example,consumers can take advantage of existing technologies (1) to alertthem when a cookie is being placed on their hard drive by awebsite, or (2) to block or remove the placement of an unwantedcookie altogether. With both Netscape Navigator and InternetExplorer, the two leading Internet browsers, a consumer canchoose to have an “alert box” flash on the screen to inform himwhenever a server is trying to place a cookie on his system. Con-sumers can employ various software packages, such as NSClean,lEClean, AdSubtract, Cookie Cruncher, Cookie Jar 2.0, InternetJunkbuster Proxy, and WebWasher to filter and block cookies inaccordance with a consumers’ individual preferences.

Various consumer information is instantly available towebsites when consumers visit them. To prevent the flow of thisinformation, consumers can use “identity scrubbers” (such asAnonymizer, Crowds and Enonymous), which are tools devel-oped to allow Internet users to remain anonymous while surfingthe Internet. Consumers also can use technologies known as digitalidentity managers or preference organizers (like Digitalme or MSPassport) to better control the manner in which their informa-tion is shared, used, and maintained online.


Consumers can assert greater control over their personallyidentifiable information by using infomediaries, which typicallyare websites that act as a “go-between” for consumers who visitwebsites but want to control the personally identifiable informa-tion that is shared at each website they visit. Some infomediariesoperate by having the consumer create a detailed personal pro-file; the infomediary then negotiates with websites over the re-lease of his personal information.

Furthermore, technologies exist to enable consumers to tradetheir personal information for something the consumer findsvaluable, such as product offers and information, sweepstakesentries, or cash. An example of this kind of consensual trading ofinformation trading is called “permission marketing,” such as thatprovided by Yesmail and Brodia, in which the consumer pro-vides personal information in order to receive offers and buyinginformation on products and services in his areas of interest.

Technologies Available to Businesses to RespectConsumer Privacy

Some tools enable online businesses that collect consumerinformation to ensure respect for consumer preferences with re-spect to personally identifiable information. One area of privacyconcern is online advertisement delivery, in which massive, sophis-ticated and detailed databases of customer profiles are developedthat enable highly targeted advertising. To the extent that someadvertisement delivery services keep consumer information imper-sonal (identifying consumer profiles by number rather than byname, for example) consumers can maintain a certain level of ano-nymity and privacy protection while receiving valuable advertisinginformation. Customer relationship management software provideshighly targeted marketing based on a consumer’s buying habits.This software can be provided, however, with services and softwareto implement personal data protection, including consumer choice(opt-in or opt-out), auditing, and security. While it is up to onlinebusinesses to implement these business-to-business tools, consumers— armed with information — can demand that retailers use suchtools and respect their privacy preferences.

Avoiding Heavy-Handed Government Regulation

Consumer confidence in e-commerce will permit our digi-tal economy to continue to grow and thrive. Consumer confi-dence in e-commerce will increase as consumers learn about anduse the tools that are available to protect their privacy online tothe extent they want. Online businesses and the government canhelp by continuing to educate consumers regarding the mannerin which personally identifiable information is collected and howit is used. In addition, online businesses must continue to de-velop and post meaningful privacy policies, undertake measuresto respond to consumer concerns about privacy, and engage inmeaningful self-regulation in order to avoid heavy-handed gov-ernment regulation.

NSDI Community DemonstrationProjects: Final ReportPrepared by the Community, Federal and Business Partners ofthe NSDI Community Demonstration ProjectsEdited by Paul Dresler and Allyson Woods, U.S. Department ofthe Interior

Editors Note: Following is the Table of Contents and ExecutiveSummary for this Report. The entire report may be found on-line at http://www.fgdc.gov/nsdi/docs/cdp

TABLE OF CONTENTS

Executive SummaryOverview of Lessons LearnedOverview of the Demonstration ProjectsIndividual Community Demonstration Project Summaries

Baltimore, MarylandSummary, PowerPointDane County, Wisconsin

Summary, PowerPointGallatin County, Montana

Summary, PowerPointTijuana River Watershed, California-MexicoSummary, PowerPointTillamook County, Oregon

Summary, PowerPointUpper Susquehanna-Lackawanna Watershed, Pennsylvania

Summary, PowerPointA Synopsis of Data Sources Used to Develop Key ProductsConclusions and Recommendations

EXECUTIVE SUMMARY

The Federal Geographic Data Committee (FGDC) together withthe National Partnership for Reinventing Government and fiveFederal agencies implemented the National Spatial Data Infra-structure (NSDI) Community Demonstration Projects betweenJuly 1998 and May 2000. This collaborative effort demonstratedthe utility of geographic data for community decision making,and highlighted the important Federal role in ensuring coordi-nation and guaranteeing access to data resources.

The Projects were designed to:• Show how cross-government, cross-functional sharing of

geospatial data, maps, expertise, and applications help solvecommunity problems;

• Support results-driven management practices using timelygeographic data;

• Strengthen efforts to set and implement cross-government,interoperable standards for data sharing;

• Supply federal expertise to communities for resolving data,policy, standards, and technical issues related to cross-gov-ernment information sharing; and,


• Share results of these pilots nationwide.

Six communities, reflecting a diversity of geographic areasand community issues were selected for the DemonstrationProjects:• Baltimore, Maryland (crime prevention and analysis);• Dane County, Wisconsin (comprehensive land use planning);• Gallatin County, Montana (Smart Growth);• Tillamook County, Oregon (flood mitigation and restora-

tion);• Tijuana River Watershed, San Diego, California (environ-

mental restoration); and the• Upper Susquehanna-Lackawanna Watershed, Pennsylvania

(flood mitigation and environmental management).

Businesses, local, state and federal agencies partnered witheach of the communities. Each community was paired with aFederal “host” agency and a point-of-contact, “Federal Cham-pion”, was identified to work with the community to facilitatetheir efforts. Additionally, a “Local Champion” was identifiedthrough a community selection process to help lead the commu-nity effort. The Project’s business partner, Environmental Sys-tems Research Institute, Inc. (ESRI), provided each communityGeographic Information Systems (GIS) software, training and asupport network to resolve technical software and data issues.

The NSDI Community Demonstration Projects demon-strated that partnerships between Federal agencies and local com-munities help facilitate the use and application of geospatial dataand tools enabling communities to make more informed deci-sions. Through such partnerships, Federal agencies received userfeedback on needed enhancements and future directions for theNSDI and Federal data and information resources. The use ofgeospatial data and visualization tools helps identify competingresources, design planning options, devlop risk mitigation strat-egies and facilitate long-term planning. It provides the capacityto inform and underpin community decision-making and provesto be an effective means to engage decision-makers, stakehold-ers, and the public at large in community planning efforts.

The Projects showed that:• Cross-government information sharing is possible without

substantial new expenditures while significantly reducing du-plication of data collection efforts.

• Mapped information is an effective means to convey infor-mation to community decision-makers and the public. Byproviding communities with a common foundation on whichto discuss planning strategies, agreement and resolution ofissues are reached in a shorter amount of time.

• When NSDI standards were applied to geospatial data, theyenhanced the ability of the user to integrate data from mul-tiple sources and incorporate the data into applications. GISis an effective tool to allow decision-makers and the publicto develop a collective, visual view of problem areas, pos-sible solutions, and a baseline to measure results in the con-text of environmental, social, and economic constraints.

• Geospatial data collected by the Federal government is per-tinent to local community issues and may, when integratedwith local data sources, be used to address a variety of localissues and problems including, but not limited to, compre-hensive planning, flooding and environmental restoration,and crime analysis as demonstrated through these Projects.

• Information required to address very localized issues such asgrowth, flooding, and crime analysis often require high resolu-tion data than is presently collected by the Federal community.

• New applications for existing data were discovered throughthe Demonstration Projects thereby increasing the useful-ness of government data acquisitions.

• In many cases a community’s existing data resources wereenhanced by NSDI standards, a result of incorporating theirdata sets with NSDI compliant databases. This added valueand additional uses to their data and facilitated the leverag-ing of resources.

• Community needs for technical assistance, data analysis, anddata clearinghouse operation can be effectively met whencommunities partner with local colleges and universities.

• Building effective cross-government data and informationpartnerships provides opportunity to leverage financial andtechnical resources and demonstrates efficient use of gov-ernment resources.

• The NSDI is an important instrument to facilitate the provi-sion of geospatial information for informed decision-making.However, the NSDI needs to evolve and expand to addressemerging needs for additional critical datasets of high resolu-tion, data and information standards, policies, and applica-tion practices. The NSDI must remain flexible and responsive.

These Projects continue to enable the six communities toseek solutions to their problems through the application of in-formation and continue to develop data about their communi-ties that has broader potential uses beyond initial Project issues.In each case the data and information is accessible through anNSDI-compliant search engine on the World Wide Web.

Recommendation: The NSDI community should initiateand expand projects to initiate a national infrastructure that fo-cuses on community data and information needs and eliminatesbarriers that communities face in working with the Federal gov-ernment to build place based information management systems.Specifically, agencies and organizations should:• Maintain and advance data and metadata standards use and

application;• Establish and link data libraries (data clearinghouses);• Evaluate, refine, build, and test ‘data applications’ libraries

and geodata ‘tool’ libraries;• Establish, test, and evaluate approaches to data consortia and

partnerships;• Explore opportunities to leverage broader investment part-

nerships and innovative financing approaches; and,• Continue to educate, train, and collaborate with others to

develop and advance the NSDI.


2000 E-Government SurveyPrepared by National Association of Counties and Public Tech-nology, Inc.

Editors Note: Following is the Executive Summary for this Re-port. The entire report may be found on-line at http://www.naco.org/pubs/surveys/it/2000egov.pdf

EXECUTIVE SUMMARY

In April 2000, The National Association of Counties dis-tributed an E-Government Questionnaire to all counties in thecountry that currently have functioning county governments. Thisquestionnaire was faxed to the Information Technology Direc-tors in all counties that have one and to County Administrators/County Clerks in those counties that did not. Counties were in-formed in the faxed communication that they could respond tothe questionnaire by fax or by using the on-line version availableon the NACo webpage.

Twenty-three percent of the nation’s counties completed andreturned the survey. These 714 counties represented a cross sec-tion of counties by region and by population size. Only the coun-ties in the states of Delaware (3 counties) and Massachusetts (6Counties) did not respond.

The availability of the tools of technology for county em-ployees is frequently determined by the size of the county. PCsare largely available to most employees in the bigger counties,including a high percentage where every employee has a PC, butnot so in the smaller counties. E-mail availability is not nearly aswide spread, since nearly half the counties indicate that employ-ees do not have email. Only a third of the counties say that eachdepartment has e-mail access.

The information collected indicates that many counties areworking towards the use of the Internet for both internal andexternal services to their constituents. The major obstacles arelack of knowledge and lack of funding. These reasons are high-lighted in responses to questions about whether counties are in-vestigating the use of the Internet which showed that many smallercounties are not aware of what it can do, haven’t thought about itor are just starting to think about it. A large majority of Countiesalso listed funding as the primary obstacle to the use of the Inter-net and to upgrading e-technology in their local government.

Many counties want to establish interactive websites allow-ing constituents to conduct business with the county on-line.However, very few currently have that capacity. The most com-mon goal is to make access to county records available to con-stituents. Nearly half the counties list this as their number onegoal.

Data for Science and Society: TheSecond National Conference on Sci-entific and Technical DataU.S. National Committee for CODATA, National ResearchCouncil

Editors Note: Following is the Preface for this Workshop Re-port. The entire report may be found on-line at http://books.nap.edu/html/codata_2nd/

PREFACE

Many of the major scientific challenges we face today re-quire the combined expertise from multiple disciplines. Com-plex issues such as the understanding of global climate change,the advance of biotechnology, and progress on various types ofproblems facing society can be addressed only by combining andusing data that in the past have been available to researchers inone field only.

In conjunction with several federal science agencies, the U.S.National Committee for CODATA organized the Second Na-tional Conference on Scientific and Technical Data: Data forScience and Society to address important multidisciplinary is-sues in managing and using scientific and technical (S&T) dataand to improve the visibility of these issues nationally. The mainfocus was on promoting the availability and usefulness of S&Tdata to all users, both in research and in the broader society, us-ing examples of ground-breaking and innovative applications andhighly creative partnerships. Three main challenges were addressedin this context:1. How can access to and use of S&T data for interdisciplinary

basic and applied research be improved?2. How can access to and use of S&T data by other sectors and

applications areas outside research (e.g., in business, educa-tion, media/entertainment, general public understanding)be improved?

3. How do we measure and evaluate productivity and perfor-mance in the management and use of S&T data within dis-ciplines, across disciplines, and in other sectors andapplications areas?

The conference was held on March 13-14, 2000, at theNational Academy of Sciences in Washington, D.C., and includeda set of plenary presentations by invited speakers, as well as con-tributed poster presentations and technical demonstrations. Theseproceedings from the conference include only the invited ple-nary presentations. These presentations have been edited and re-viewed according to standard National Research Councilprocedures, although they are included in this volume as the con-tributions of each individual speaker.


Declaring Independence: A Guide toCreating Community-Controlled Sci-ence JournalsThe Scholarly Publishing and Academic Resources Coalition(SPARC)

Editors Note: Following is the Table of Contents and Introduc-tion for this Report. The entire report may be found on-line athttp://www.arl.org/sparc/DI/

TABLE OF CONTENTS

Table of Contents1 Dear Colleague Michael L. Rosenzweig, Editor, Evo-

lutionary Ecology Research

2 Introduction Rick Johnson, Enterprise Director,SPARC

3 Stage One Diagnosis

6 Stage Two Exploring Alternative Options

10 Stage Three Evaluating the Options

14 Conclusion Making the Decision

15 Appendix A Library Journal 2000 Periodical PriceSurvey

16 Appendix B Web Resources

18 Appendix C Selected Bibliography

INTRODUCTION

The Data. During the last 4 years, the average cost of a com-mercially published scientific Journal has risen nearly 50%.

The Effect. College and research libraries unable to keeppace with these rising costs are canceling journals and divertingincreasing amounts of the acquisitions budget to cover the costof those remaining.

In the publications explosion of the past few decades, com-mercial firms have found there was a profit to be made in thevaluable service of publishing research journals. Scholars, in needof promotion and tenure, were happy to publish in the commer-cial journals-especially when then the alternative was not beingpublished at all-and gave their research papers away to journalsfree of charge.

A few commercial publishers discovered that the easiest wayto increase profits was to raise subscription prices and, moreover,that the fattest profits came from raising library subscription pricesaggressively and relentlessly. Institutional subscribers accounting

for the lion’s share of the revenue supporting publication of jour-nals in most fields, paid the price-reluctantly and with increasingdifficulty-because their users demanded access.

With this foot in the door, these few commercial publishersbuilt substantial portfolios of journals, aided by the trend of schol-arly societies outsourcing their journals to commercial firms. Thehigh profits from these journals have funded wave upon wave ofacquisitions and consolidations among publishers, and often theseprofits are diverted out of scientific activities into unrelated linesof business in order to “enhance shareholder value.”

The results are clear: high prices, declining circulation, inef-ficient production schedules, and in many cases, lack of respon-siveness to editors, editorial boards, and authors. But you andyour colleagues can change this.

As you know every experiment derives from a hypothesis.SPARC’s hypothesis is that high impact, low-cost scientificjournals-published by societies, university presses, or indepen-dent publishers-can provide researchers with prestigous alterna-tives to expensive commercial journals. SPARC partners haveproven that with high quality and reasonably priced alternativesto journals that no longer serve the community of scholars well.At the same time, SPARC members pledge support for SPARCsponsored publications. This winning formula has given scien-tists new, responsible outlets for publishing their research and agood chance of financial visibility over the long term.

Declaring Independence is your guide to addressing the sci-entific communication crisis. It provides tools for evaluating yourpresent situation and for evaluating publishing alternatives, aswell as an approach to assessing those options.

Only you can determine the path that best suits you andyour field. If you decide that Declaring Independence is yourpath, SPARC is here to support you.

Rick JohnsonSPARC Enterprise Director

All applications will be reviewed by the ESIG™ Committee and winners will be notified in July, 2001. Winners will be recog-nized during the Awards Ceremony at URISA 2001 in Long Beach, and one person from each winning system willreceive a complimentary full registration for the conference. Following the conference, winners will receive addi-tional recognition in URISA publications and an announcement of their accomplishment will be made to media representativesaround the world.

In order for the ESIG™ Review Team to fairly evaluate each system, specific information (A-F below) must be included in yoursubmission. If submitting this application in electronic form (preferred), send it as an email attachment to [email protected] in PDFor Microsoft Word format. In the body of the email, specify the format, version number, and the length of the attached docu-ment. Include “ESIG™ Application” in the email subject field. If submitting in paper form, please send twelve (12) copies to:

URISAESIG™ Application1460 Renaissance Drive, Suite 305Park Ridge, IL 60068-1348

Provide all requested information in your submission. Incomplete applications will not be considered.The application deadline is: June 1, 2001

A. System1. Name of system and ESIG™ category for which you are applying (Enterprise System or Single Process System).

ESIG™ Award Categories:Enterprise Systems: Systems in this category are outstanding and working examples of using information systems technologyin a multi-department environment as part of an integrated process. These systems exemplify effective use of technology yieldingwidespread improvements in the process(es) and/or service(s) involved and/or cost savings to the organization.

Single Process Systems: Systems in this category are outstanding and working examples of applying information system technol-ogy to automate a specific SINGLE process or operation involving one department or sub-unit of an agency. The system applicationresults in extended and/or improved government services that are more efficient and/or save money.

2. A letter from the executive administrator authorizing submission of the system application. (Include as a separate attachment ifsubmitting electronically.)

3. One (1) page, or less, summary of what the system accomplishes and why it is exemplary.4. Three “user testimonials”. These testimonials should include the title of the system, the person’s name, job title (if relevant), a

statement of what specific ways the system improves their work and/or the work of their organization, and how frequently theyuse the system. (Include as separate attachments if submitting electronically.)

B. Jurisdiction1. Name of jurisdiction2. Population served by the organization/agency3. Annual total budget for jurisdiction4. Name, title, and address of chief elected and/or appointed official5. Name, title, address, telephone, FAX, and email for contact person for system

Apply for a URISA 2001 ESIG™ Award!

If so, that achievement should be recognized

and shared with your peers. Nominate your

organization for a prestigious URISA Exemplary

Systems in Government (ESIG™) Award. Or

convince a colleague to participate!

Has your organization improved the delivery

and quality of government services through

the application of information technology?

You must answer each of the following questions. Please cross-reference your responsesto each of the topics/questions listed below. Be sure that your responses are clearly writtenand sufficiently comprehensive for reviewers to develop a clear understanding of thesystem. Responses should be in complete sentences and as brief as possible while communi-cating the necessary information. If appropriate, include graphics.

C. System Design1. What motivated the system development?2. What specific service or services was the system intended to improve?3. What, if any, unexpected benefits did you achieve?4. What system design problems were encountered?5. What differentiates this system from other similar systems?

D. Implementation1. What phases did you go through in developing the system?2. Were there any modifications to the original system design? Why? What?

E. Organizational Impact1. What user community does the system serve and how?2. What are the ultimate decisions/operations/services being affected? If appropriate, provide a few examples including, but not

limited to: screen input/output forms, paper products, or other descriptive graphics.3. What were the quantitative and qualitative impacts of the system?4. What effect has the system had on productivity?5. What, if any, other impacts has the system had?6. How did the system change the way business is conducted with and/or service delivered to clients? Give specific examples

comparing the old way with the new.

F. System Resources1. What are the system’s primary hardware components? Give a brief list or description of the hardware configuration supporting

the system.2. What are the system’s primary software components? Describe the primary software and, if a commercial package, any

customizations required for the system.3. What data does the system work with? List and briefly describe the database(s).4. What staff resources were required to implement the system (i.e., report approximate staff and consultant time as FTE’s)

Join the exclusive list of ESIG™ Award

winners. If you’ve successfully improved

the way in which government operates,

through the use of information technology,

you should apply for a 2001 URISA

ESIG™ Award. If you have any questions, contact URISA

Headquarters at (847) 824-6300 or

[email protected]

Ap p l i c a t i o n D e a d l i n e : June 1, 2 0 01

One person from each

winning system will receive

a complimentary registra-

tion for URISA 2001!

eMapsPlus is made possible by a national network of data providers from both thepublic and private sectors who are committed to making the most up-to-dategeographic content available online to subscribers.

eMapsPlus is the most advanced online service available for accessing GISproperty data. This powerful tool is perfect for the specific needs of the casualuser as well as the wider, on-going commercial requirements of property relatedbusinesses. eMapsPlus allows immediate Internet access to the advancedinformation and resources of SDS -- an industry leader in GIS property datamaintenance.

Discover the many benefits of eMapsPlus . . . Whether you require rich datasetsspanning several counties or just the deeded acreage of one specific parcel,eMapsPlus allows you to easily navigate to the information you need. You canorder data as ESRI shapefiles, ARCInfo coverages, or in the popular DXF fileformat.

And, if you're a data owner, SDS will manage and market your GIS propertyinformation for third-party usage creating a new, trouble-free revenue source.

SDS announc e s

The online, ONE STOP DATAWAREHOUSE for all yourproperty information needs.

SDSSmart Data Strategies357 Riverside DriveFranklin, Tennessee 37064

T 615.794.5280F 615.794.5310corporate 888.384.6214www.sds-inc.com

Stop by our eMapsPlus warehouse today atwww.eMapsPlus.com

To learn more about our flexible data hosting plans call oursales department at > 1.888.384.6214

To _____________________________________ Company _______________________________________

Date ___________________________ Job No. __________________________________________________

Job Description ____________________________________________________________________________

No. of Pages (Including Cover Sheet) _________

Comments:

You are being sent the following:Date Sent (JL Design) Date Returned (Client)

❑ First draft ____________ ____________

❑ Draft No. 2 with revisions ____________ ____________





If this is a final proof of your job, carefully review and sign off. NOTE: You should re-reviewall copy before approving—not just last round of revisions.

❑ Please make the following noted revisions before approval. (Do not sign below if changesneed to be made.)

❑ Proof is OK—please go to print.I have examined this proof for spelling, color breaks, photos, and all other elements Irequested. I understand that any errors found at a later time are my responsibility.

For Approved Final ProofsI have reviewed the final proof of my job and approve it.

Signed __________________________________________________ Date __________

TYP E S E TT ING APPROVAL FORM

When reviewingtypesetting, it isimportant tocheck for thefollowing:

� Layout anddesign matchspecs

� Informationthat wassupplied is allincluded

� Spellingand grammarare correct

� All re-quested editshave beenmade

� Colorbreaks are setas you wantthem

� Pictureshave correctcaptions andare cropped toyour specs

183 E.Regent Dr.Clarksville, TN37043

phone931.358.0533

fax931.358.0791

Scott URISA

Feb 19, 2001 jl01-022

winter 2001 journal

feb 19 (PDF) 3p March 1(fax)

March 5 (PDF) 1p March 6 (email)

March 7 (PDF) 10a March 8 (email)




March 19 (PDF) 5:30p

Documents

Volume 13 • Number 1 • Winter 2001 Journal of the Urban and Regional …€¦ · · 2016-05-12Journal of the Urban and Regional Information Systems Association CONTENTS