EIS Executive Information Systems, Inc

1

1

Enterprise Knowledge Portals In eBusinessPrepared By:

Joseph M. Firestone, Ph.D.Chief Scientist

Executive Information Systems, Inchttp://www.dkms.com

[email protected](703) 461-8823

February 20, 2001

Prepared For:Enterprise Intelligence Conference

Executive Information Systems, Inc.EIS

© 2000 Executive Information Systems, Inc.

2

2

IT Applications Are PurposefulgThey Support and Partially Automate

Human Participation In BusinessProcesses

gPortals are No DifferentgSo Let’s Talk About eBusiness and

Knowledge Processes Before We TalkAbout Knowledge Portals


3

3

eBusiness


g eBusiness is not the same as eCommerceg eCommerce is about selling over the Internet, whileebusiness is about using web technology to support a setof related business processes including Salesg Examples of other e-business processes are:

g eCRM (web-enabled Customer RelationshipManagement),g eSCM (web-enabled Supply Chain Management),g eERP (web-enabled Enterprise Resource Planning),g eKP (web-enabled Knowledge Processing, andg eKM (web-enabled Knowledge Management)

4

4

g Customg EIP Technology

gEIPsgeIPs

g EKP TechnologygEKPsgeKPs

Types of Web Enablement in eBusiness


5

5

What is anEnterprise Information Portal?

gAccording to Merrill Lynch's Shilakes andTylman:

g "Enterprise Information Portals are applications thatenable companies to unlock internally and externallystored information, and provide users a singlegateway to personalized information needed tomake informed business decisions."


6

6

What is an EnterpriseInformation Portal? (TWO)

g ". . . an amalgamation of software applicationsthat consolidate, manage, analyze and distributeinformation across and outside of an enterprise(including Business Intelligence, ContentManagement, Data Warehouse & Mart and DataManagement applications)."


7

7

What’s An eIP?

g An eIP is an ebusiness Information Portal, thatis, it uses EIP technology to support e-businessprocesses that transcend the enterprise.

g There are two basic types of eIPs:gExtraprise Information Portals (ExIPs); andg Interprise information Portals (IIPs)


8

8

Extraprise Information Portals

g An ExIP is an Information portal supporting anextended enterprise usually consisting of acommunity of trading partners having a mutualinterest, revolving around a common hostenterprise, who do business with one anotheron a fairly predictable and repetitive basis. Theenterprise at the center of the system usuallyhosts the “extended Intranet” (aka, the“Extranet”).


9

9

Interprise Information Portals

g An Interprise Information Portal (IIP) is anInformation portal supporting web-likefederations of otherwise independentcompanies with no “network host” at the center.

g The members of the interprise do businesswith one another through the IIP on a fairlyunpredictable and irregular basis in responseto individual expressions of demand inmarketplaces of mutual interest.


10

10

Problems andSolutions in the Enterprise

g Every individual, team, or group within the enterpriseencounters problems in the course of the work day as theyattempt to make decisions. A problem is created by anattempt to make a decision followed by a realization that onecan’t make it because one can’t predict the outcome

g Every problem has alternative solutionsg And every alternative solution is subject to criticism and to

replacement if it performs less well than its competitors.g The best problem solution is the competitive alternative that

best survives criticism© 2000 Executive Information Systems, Inc.

11

11

The Enterprise is a “Swirl” ofKnowledge-related Interactions

g The set of problem-solving interactions in anenterprise constitutes a continuous, dynamic “swirl”from which knowledge is produced and integratedwith the business processes of the enterprise

g For a given problem, it is useful to abstract from theswirl and to conceptualize an iteration of a knowledgelife cycle targeted on solving that problem.


12

12

The Knowledge Life Cycle (KLC) Model


KnowledgeProduction

OK

InfoAboutVKC

VKC

InfoAboutUKC

UKC

InfoAbout

IKC

IKC

KnowledgeIntegration

Knowledge Processes

Knowledge Sets

UKC - Unvalidated Knowledge ClaimVKC - Validated Knowledge Claim

IKC - Invalidated Knowledge Claim

OK - Organizational Knowledge

ExternalInputs

Experiential Feedback Loop

Distributed Organizational

KnowledgeBase

Structures incorporating Organizational

Knowledge

Bus.ProcessBehavior

ExternalInputs

13

13

Structures IncorporatingOrganizational Knowledge

gg Business ProcessesBusiness Processes

gg Organizational CultureOrganizational Culturegg Organizational StrategyOrganizational Strategygg Organizational TeamsOrganizational Teamsgg Formal Org. Sub-divisionsFormal Org. Sub-divisionsgg IndividualsIndividualsgg PoliciesPoliciesgg ProceduresProceduresg Products

g Servicesg Codified Organizational

Knowledgeg Information Systemsg Paper documentsg Imagesg Artg Other Organizational

Cultural Artifacts


14

14

Knowledge Production


CKC

InfoAcquisition

Indiv. &Group

Learning

CKC - CodifiedKnowledge Claims

KnowledgeValidation

KnowledgeClaim

Formulation

15

15

KnowledgeSharing:

Face-To-Face,Document,Computer-

based

OK

Searching/Retrieving:Electronic

or Personal

Teaching:Face-To-Faceand Compu-

ter based

Broadcasting:Electronic

or Personal

Knowledge Integration


16

16

What is Knowledge Management?

g It is handling, directing, governing, controlling,coordinating, planning, and organizing agents,components, and activities participating in the basicknowledge processes (knowledge production andknowledge integration)

g That is, it is managing the KLC -- its processes and itsproducts (outcomes)

g Such management occurs through a range of activities


17

17

KM Task Clusters and Task Patternsg There are three KM task clusters: interpersonal

behavior; knowledge processing behavior; and decisionmakingg Interpersonal behavior includes three task patterns:

gfigurehead,gleadership, andgexternal relationship-building activity.

g Knowledge processing behavior includes two taskpatterns:gKM knowledge production; andgKM knowledge integration.


18

18

KM Task Clusters andTask Patterns (TWO)

g Decision Making includes four task patterns:g changing knowledge process rules;g crisis handling;g allocating KM and knowledge processing

resources; andg negotiating agreement with representatives of

other business processes.


19

19

g An EKP is an enhanced Enterprise InformationPortal (EIP)

g It is an EIP that supports knowledge production,knowledge integration, and knowledgemanagement

g It is an EIP that supports individuals, groups, andteams in the swirl of problem-solving activitiespermeating enterprise business processes

What is an EnterpriseKnowledge Portal (EKP)?


20

20

g focus upon, provide, produce and integrate informationabout the validity of the information they supply

g provide information about your business and meta-information about the degree to which you can rely onthat information,

g distinguish knowledge from mere information,g provide a facility for producing knowledge from

informationg orient one toward producing and integrating knowledge

rather than information

Enterprise Knowledge Portals


21

21

Portal Browser Clients

DataMarts

Content and Data Stores

DDS DW ODS PersistentContent Store

(OODBMS)

OLTPRelational

Full-textIndexed

ApplicationServers

Collab

Text and Data Mining Server

Web Server

Portal JobContent Server

PageBuilder

Transaction &Report/Query

Scanner/Indexer

Agent Server

ETL, ROLAP &MOLAP

ERP &Legacy

ApplicationServers

Agent Agent

Agent

ArtificialKnowledge

Server

Artificial Knowledge Manager

The Architecture of an Enterprise Knowledge Portal


22

22

EKP Generic Application Componentsg Browser and e-mail clientsg The Avatar -- a client-based intelligent agentg The portal application server(s),g The access management systemg Knowledge Claim Objectsg The enterprise Artificial Knowledge Server(s) (AKSs),g Complex adaptive system (cas) mobile intelligent agentsg The formal knowledge production application server(s) and its

associated clients supporting analytical and statisticalmodeling, KDD and Data Mining, Simulation, impact analysisand forecasting,

g The collaborative processing application server, andg A persistent storage component.


23

23

EKP User Interface: Start-up and the Avatar

g The EKP provides a personalized user interface for each knowledgeworker to use the portal.

g The organization of this interface is at first determined by an initialprofile of desires and interests determined by a knowledge workerdialog with the portal upon first use. This dialog serves to program anintelligent agent serving the portal interface.

g On initial programming, the intelligent agent becomes the knowledgeworker’s avatar, representing it to the portal system.

g Thereafter, the avatar will modify the portal’s content and organizationbased on preferences expressed in actual patterns of use of the portalby the knowledge worker.


24

24

EKP User Interface: The Avatar

g The avatar uses these patterns of use to learn the knowledge worker'sinterests and preferences over time.

g The avatar also learns the changes in the knowledge worker's interestsand preferences. Eventually the avatar uniquely represents theknowledge worker and ensures that the EKP is entirely personalized tothe knowledge-centric work flow patterns of the knowledge worker.

g The EKP then becomes a desktop environment keyed to the knowledgeworker's cognitive map and work flow and the "islands of automation"characteristic of today's desktops are replaced by desktops supportingintegrated work flows defined in the context of the knowledge worker’scognitive map.


25

25

What is the Purpose of the Avatar in the EKP?

g Creating a highly automated, enterprise knowledge-enabled, self-learning/adaptive interface that ispersonally and dynamically tailored to each user. Thetypes of automation involved here include:g assimilation of the user's local environment, personal

preferences & cognitive patternsg learning from, interacting with, and utilizing enterprise

knowledge communicated by the Artificial KnowledgeServers


26

26

What is the Purpose ofthe Avatar in the EKP? (Two)

g supporting personal and collaborative workflows, by providingworkflows retained in memory, and accessible from the portalinterface

g adapting according to anticipated needs of users, determinedthrough the Avatar’s access to user cognitive maps

g producing knowledge claims and submitting these to the EKPsystem. These knowledge claims of the Avatar representlocal knowledge, in contrast to the rules that have beenvalidated by the EKP system’s network of Artificial KnowledgeServers

g providing the user with autonomous negotiating capabilitieswith the rest of the EKP system


27

27

Knowledge Claim Objectsg An important class of objects in the EKP system is the knowledge

claim object (KCO)g A KCO is distinguished from an ordinary business object by the

presence of validity metadata encapsulated in the objectg Such metadata compares the KCO to alternative, competing

KCO’s, and may be expressed in many different forms. The“metadata” may be qualitative or quantitative or it may be in theform of textual content. In relatively infrequent but importantspecial cases, the metadata may involve quantitative ratings of aknowledge claim compared to its competitors.

g When the KCO is accessed by a user, data, metadata, andmethods are all available, so the user can evaluate the KCO as abasis for decision against competing KCOs. This capability is notavailable in EIPs, which express knowledge claims as data orbusiness objects only.


28

28

Process Control Services

Active In-Memory Object Model

Connectivity Services

The AKM


Process Control Services:

In-memory proactive object state management and synchronization across distributedobjects

Component management and Workflow Management

Transactional multithreading

business rule management and processing, and

metadata management.In-memory Active Object Model/Persistent Object Store is characterized by:

Event-driven behavior

EKP-wide model with shared representation

Declarative business rules

Caching along with partial instantiation of objects

A Persistent Object Store for the AKS and

Reflexive ObjectsConnectivity Services are:

Language APIs: C, C++, Java, CORBA, DCOM

Databases: Relational, ODBC, OODBMS, hierarchical, network, flat file, etc.

Wrapper connectivity for application software: custom, CORBA, or COM-based.

Applications connectivity including all the categories mentioned above, whether these aremainframe, server, or desktop - based.

29

29

The Artificial Knowledge Serverg The distributed AKS provides Process Control Services, an Object Model of

the EKP system, and connectivity to all enterprise information, data stores,and applications

g Process Control Services:g In-memory proactive object state management and synchronization

across distributed objectsg Component management and Workflow Managementg Transactional multithreadingg business rule management and processing,g KCO management and processing andg metadata management

g In-memory Active Object Model/Persistent Object Store is characterized by:g Event-driven behaviorg EKP-wide model with shared representation


30

30

The Artificial Knowledge Server (Two)g Declarative business rulesg Caching along with partial instantiation of objectsg A Persistent Object Store for the AKS andg Reflexive Objects and KCOsg Connectivity Services are:

g Language APIs: C, C++, Java, HTML, XML, CORBA, DCOMg Databases: Relational, ODBC, OODBMS, hierarchical,

network, flat file, XML, etc.g Wrapper connectivity for application software: custom,

CORBA, or COM-basedg Applications connectivity including all the categories

mentioned above, whether these are mainframe, server, ordesktop - based


31

31

The Intelligent Agent Platformg The second type of component comprising the EKP’s AKM is the intelligent

agent (IA).g EKP IAs are lightweight, intelligent, efficient, specialized Business Process

Engines that provide some memory and a small amount of processing powerat almost no cost.

g Agents alone cannot yet create the virtual enterprise. For complex processingand an enterprise wide view, the AKS is also indispensable.

g But IAs provide distributed load balancing to processing in the AKM.g They are necessary partners in providing the processing power needed for

implementing the EKP as a virtual enterprise.g When we add agents to the AKS to create the AKM, we provide software

wiring for the enterprise that connects its central brain components (the AKSs)to its sensors (the agents). The result is a flexible and scalable AKM that canintegrate the various components of the EKP into a virtual enterprise.


32

32

The Intelligent Agent Platform

g Lightweight expert system w/rule-based inference engine supportingforward/backward chaining and ‘what if’ scenarios

g Lightweight high-performance multi-threaded SNMP engine thatprovides two-way communication between agents and othercomponents of the EKP system

g Component for modeling cycle time, blockages in a system, work flows,and interpersonal, and semantic networks

g Complex Adaptive Systems agents (agents interact with localenvironment and external components to automatically formulate localknowledge and submit knowledge claims to the next higher level in thesystem hierarchy which automatically tests and validates them. Thisprocess produces automated knowledge production and automatedadaptation to local and global environments)


33

33

g Competitive Advantageg Increased ROIg Increased Employee Productivityg Increased Effectivenessg Decreased Cost of Informationg Increased Collaborationg Universal Access to Enterprise Resourcesg A Unified, Dynamically Integrated and Maintained

View of Enterprise Data and Informationg Accelerated Innovation

Some Claimed Benefits of EIPs


34

34

g While these are benefits claimed for Enterprise InformationPortals:g Competitive Advantageg Increased ROI,g accelerated innovation andg Increased Effectiveness -- the four most important benefits

g tacitly assume that the information produced or supplied by anEIP is correct information

g Risk: If that’s not the case, these four benefits are lostg An overriding justification for implementing an EKP, rather than

an EIP, is to secure these four benefits and to minimizedecision making risk by increasing the quality/validity ofinformation supplied by the portal

Benefits of EKPs


35

35

g An eKP uses EKP technology to support e-business processes that transcend theenterprise.

g There are two basic types of eKPs:gExtraprise Knowledge Portals (ExKPs); andg Interprise Knowledge Portals (IKPs)

What’s an eKP?


36

36

g The use of EIPs in eBusiness is an acceleratingtrend

g EKPs provides all the support for eBusinessprovided by EIPs. In addition, EKPs provideunique support for:

g eCRMg eSCMg eERPgeCommerce

The EKP/eKP and e-Business


37

37

The KLC Model and eBusiness


Distributed Organizational

Knowledge Base

eERP

eSCM

eCRM

eKP

eKM

eCommerce

Structures Incorporating

Organizational Knowledge

External Inputs

38

38

g Knowledge-based Personalization of CRM webclients

g Customer individual and group learningg Knowledge worker individual and group learningg Knowledge validation in such areas as:

g Strategic and Tactical CRM planningg Customer acquisitiong Customer retentiong Customer lifetime Valueg Monitoring and evaluating CRM initiatives

The EKP/eKP and eCRM


39

39

g Knowledge-based Personalization of SCM webclients

g Knowledge worker individual and group learningg Knowledge validation in such areas as:

g SCM planning and process modelingg Raw material developmentg Ingredient forecastingg Manufacturing process controlg Inventory forecasting and control

The EKP/eKP and eSCM


40

40

g Knowledge-based Personalization of ERP webclients


g Budgetingg Accountingg Asset Managementg Human Resourcesg Shipments

The EKP/eKP and eERP


41

41

g Knowledge-based Personalization ofeCommerce web clients


g eCommerce planningg Sales forecastingg Billingg Collectiong Ordersg Deliveries

The EKP/eKP and eCommerce


42

42

g Though there are vendors who claim to sell EKPs(IBM/Lotus, Hyperwave, Practicity, Comintell, Unisys,etc.), there are no EKPs or eKPs yet.

g There are no portal products that focus on a concept ofthe knowledge life cycle and on a specific set of activitiesthat are recognizably KM activities. And there are noportal solutions that do this either.

g Instead, products and solutions that carry the name"knowledge portal" seem to use it for its marketing valuerather than because it is descriptive of the businessprocesses that the portal product or solution supports.

The EKP/eKP Is A Vision, Not a Reality


43

43

g Once again, because competitive advantage, increasedROI, accelerated innovation and Increased effectiveness-- the four most important benefits claimed for EIPs --tacitly assume that the information produced or suppliedby an EIP is correct information

g If that is not the case, these four benefits are lostg So, to implement an EIP is to place these benefits at risk.g If we are to seriously pursue them, we must implement

the EKP to minimize decision making risk by increasingquality/validity of information supplied by the portal

Why Seek the Vision


44

44

The End


45

45

g Decision Processing (Viador, CA, Brio One)g Content Management (Plumtree, Autonomy, Verity,

Oracle, IBM, KnowledgeTrack, SageMaker,DataChannel, Sequoia Software)

g Collaborative Processing (Practicity, Engenia,Intraspect, OpenText)

g Decision Processing/Content Management(Hummingbird, Sybase, Iona)

Portal Segments


46

46

g Advanced Collaborative Processing (DP/CM +Collaborative + some Knowledge Production + additionalfeatures)

g Structured Information Management (ACP minus MostCM features of ACP)

g Structured Knowledge Processing (ACP minus- Most CMfeatures of ACP + Knowledge Processing and KM)

g Comprehensive Knowledge Processing (ACP +Knowledge Processing and KM)

More Portal Segments


47

47

g Begin with a contemporary portal type, e.g., Content Management,Decision Processing, Collaborative, or Decision Processing/ContentManagement portal

g Construct the portal solution using an architecture that includes:g Browser and e-mail clientsg The portal and other necessary application server(s),g The access management systemg An enterprise Artificial Information Manager (AIM), andg A persistent storage component.

g Add Application Servers, the Intelligent Agent Components, andKnowledge Claim Objects integrating them using the AIM

How to Seek the Vision:The Incremental Road to the EKP


48

48

DP IM SKn

CM DP/CM CompKn

AdvCollab

Collab

Pathways of Portal Evolution


49

49

eCommerce eCRM eSCM eERP

EKPs and eKPs Support andIntegrate eBusiness ProcessesEKPs and eKPs Support and

Integrate eBusiness Processes© 2000 Executive Information Systems, Inc.

50

50

Object-based Integration/Synchronization Server


Browser Clients

Data MartsODS OLTP

RelationalApplication Servers

Exported Services

Operational Object Model

Intermediate Classes

InterlayerMappings

ObjectModel with

EventHandling and

InferenceEngine

51

51

ModelSpecification

MeasurementModel

A Cognitive Map

DefinitionDefinition

Measures

Measures

AbstractionsBoundary

Boundary© 2000 Executive Information Systems, Inc.

Documents

EIS Executive Information Systems, Inc