1
Semantic WebVision, Technologies,
ApplicationsProf. Eero Hyvönen
Helsingin yliopisto ja Tietotekniikan tutkimuslaitos HIITSemantic Computing Research Group
http://cs.helsinki.fi/group/seco/
Tekoäly, Eero Hyvönen, 2004
2
Outline
• WWW as publication channel– Where are we now?
• Semantic web– Ontologies constitute the core
• Applications & demos
2
Tekoäly, Eero Hyvönen, 2004
3
Ontologies – Perspectives
– Philosophy– Linguistics– Computer Science
Tekoäly, Eero Hyvönen, 2004
4
Perspectives of ”ontology”:Philosophy
• Study of the essence of Being – apart from the particular existing things
• Plato’s world of ideas: metaphysics• Artistotle’s (384-322 B.C.) 10 Categories• Medieval logicians: first semantic net
– Genus (supertype) vs. species (subtype)• “Ontology” as a discipline
• R. Göckel, J. Lorhard, 1613• Kant (1787), Peirce, Husserl, Whitehead,
Heidegger, …
• Today: theoretical bias– foundational categories & logic behind
everything
3
Tekoäly, Eero Hyvönen, 2004
5
Aristotles’ 10 categories
The cat satPositionThe cat has a ratHavingThe cat is runningActionThe cat desires fishPassion
The cat came out yesterdayWhenThe cat is in the houseWhereThe cat is half the size of …RelationThe cat is 50cm highQantityThe cat is blackQualityA catSubstance
Tekoäly, Eero Hyvönen, 2004
6
Hierarchical categories: Tree of Porphyry of Aristotle’s ”Substance”
4
Tekoäly, Eero Hyvönen, 2004
7
Aristotle’s Syllogisms• Four types of
propositions:
• Examples of syllogisms:
(Sowa, 2004)
Tekoäly, Eero Hyvönen, 2004
8
Formal ontology• Branch of philosophy
– Using formal methods in the study of being– Developing formal (logical) ontological theories– Combination of philosophy and AI
• Theories– Theory of parts & wholes– Theory of time– Naiive physics– …
• Domain dependent vs. independent ontologies• Descriptive vs. formal ontologies
5
Tekoäly, Eero Hyvönen, 2004
9
Perspectives of ”ontology”:Linguistics
• Meanings vs. words• Peter Mark Roget’s Thesaurus 1852
– Tool for analysis and classification of ideas that helps human communication
• Terminology: dictionaries & vocabularies– Concept analysis
• Thesauri– Indexing/classifying/retrieving data– Language translation
• Semantic thesauri– WordNet, EDR (Electronic Dictionary
Research),…
Tekoäly, Eero Hyvönen, 2004
10
Roget’s Thesaurus
• Everything in 1000 categories– Nouns, adjectives, verbs, …
• 1975: over 100,000 words– Neighboring categories semantically related
• e.g. 266=”Journey”; 267=”Navigation”– Not a formal model
• Only for human interpretation
6
Tekoäly, Eero Hyvönen, 2004
11
Roget’s Thesaurus: example• Top level: 6 classes
• 1. Abstract relations• 2. Space• 3. Matter• 4. Intellect• 5. Volition• 6. Affections
• CLASS 2. SpaceI Space in general
Abstract space180 Indefinite space {Noun: space, extension, extent, expanse,…
Verb: reach, extend,…Adj: spacious, roomy, …Adv: extensively, …}
181 Definite region …182 Limited space …
II Relative space183 Situation ……
Tekoäly, Eero Hyvönen, 2004
12
Terminology
• Based on concept analysis of word meanings• Standardized methodology
– Various ISO standards• Well-defined vocabularies for human users, not
for machines– In Finland: e.g. Sanastokeskus TSK
• Following presentation is based on:– Heidi Suonuuti: “Guide to terminology”, 2001
7
Tekoäly, Eero Hyvönen, 2004
13
Concept analysis:Extended Odgen-Richards triangel to tedraed
Object
Definition
Terms
Concept
treeBaumarbrepuu
”tall plant with hard self-supporting trunck and …”
Tekoäly, Eero Hyvönen, 2004
14
Concept vs. term• Monosemy
– one term - one concept• Polysemy
– one term - many related concepts• E.g. “head” (of arrow) vs. “head” (of human)
• Homonymy– one term - many unrelated concepts
• E.g. “kuusi” (number vs. tree)
• Synonymy– One concept – many terms
• E.g. “honka”=“mänty”
8
Tekoäly, Eero Hyvönen, 2004
15
• Characteristics (properties)– Delimiting characteristics differentiate concepts
• E.g. concept “tree”: – “have a root” not delimiting– “have a self-supporting trunk” delimiting
• Intensional and extensional definitions– Intension = sum of general characteristics
• Tree= “long-living plant, have a self-supporting trunk, …”
– Extensional = list of objects• Tree= {pine, maple, spruce, …}
Tekoäly, Eero Hyvönen, 2004
16
Concept systems
• Concepts are not independent• Relations between concepts
– Generic (hyponyny)– Partitive (meronymy) – Associative (function, et.)
9
Tekoäly, Eero Hyvönen, 2004
17
Generic relation (hyponymy)
• Concepts share general characteristics but one• Concept hierarchy: Super/subordinate
• Problem: several branching possibilities– Anatomy: coniferous vs. broad leaf– Requirements: light-demanding vs. tolerant– Abscission: evergreen vs. deciduous
tree
coniferous tree broadleaf tree
pine spruce … birch maple …
Tekoäly, Eero Hyvönen, 2004
18
Representing parallel independent subdivisions
tree
evergreen deciduoustree tree
abscissionanatomy
requirements
coniferous broadleaftree tree …
pine spruce …
Using three subdivision dimensions:
10
Tekoäly, Eero Hyvönen, 2004
19
Partitive relation (meronymy)• Part-whole relation
– Examples:• Atoms in a molecule• Legs of a chair
– Optional, single, and multiple parts
– Also along different dimensions• Tree -> permanent vs. non-permanent organs
tree
root trunk branch
Tekoäly, Eero Hyvönen, 2004
20
Different meronymy relations
childhood/growing-upphase /processHelsinki/Finlandplace / region
aluminium/airplanematerial / objectpiece-of-cake/cakepiece / wholetree/forestmember / setbranch/treepart / whole
(C. Fellbaum, 1998)
11
Tekoäly, Eero Hyvönen, 2004
21
Associative relations
nest / treeobject / location
etc…etc…wood / papermaterial / productpaper machine / paper makingtool / functionapple tree / fruit gatheringobject / activity
nesting / treeactivity / locationnesting / birdactivity / actormagpie / nestproducer / productspring / leafs in treescause / effect
Tekoäly, Eero Hyvönen, 2004
22
Associative relations
• Arrow notation
pulping
tree wood
process / material
origin / material
12
Tekoäly, Eero Hyvönen, 2004
23
Term definition
• Based on:– Concepts in the concept system– Relations among the concepts
• Intensional definitions– Idea: describe only delimiting characteristics– Other characteristics come from the hierarchy
• “noble gas”: gas that in the natural state is inactive
• Extensional definitions: list objects• “noble gas”: helium, neon, argon, krypton, and radon
Tekoäly, Eero Hyvönen, 2004
24
• Principle of substitution– Terms and definitions are interchangeable
• Deficient definitions– Circular definitions
• Evergreen tree: – Bad: tree with evergreen foliage– Good: tree that retains its foliage throughout the year
– Negative definitions• Deciduous tree:
– Bad: other than evergreen tree– Good: tree losing its foliage annually
13
Tekoäly, Eero Hyvönen, 2004
25
– Incomplete definitions• Too broad:
– Tree: tall plant that lives for many years» Covers banana, wine, …
• Too narrow:– Fertility: ability of a tree to produce offspring
» Also magpies can be fertile
– Hidden definitions• Several definitions in a concept definition
Tekoäly, Eero Hyvönen, 2004
26
Terms
• Term types– Word
• Nouns, adjectives, verbs• “house”, “biodegradable”, “reclaim”
– Compound• “lighthouse”
– Multiword expressions• “broadband integrated services digital network”• ”kolmivaihekilowattituntimittari”
14
Tekoäly, Eero Hyvönen, 2004
27
• Preferred terms– Usually only one preferred term per concept– Other synonyms: admitted / deprecated
• Selection of terms– Concepts in a subject field– Concepts used in several subject fields– Concepts borrowed from adjacent fields– Concepts used in general language
• Harmonization– Identical terms between different fields
Tekoäly, Eero Hyvönen, 2004
28
Why terminology
• Provides useful tools and ideas for concept analysis and definition
• Normative goal– Analyze, select, harmonize, and define a concise set
of terms to be used in human communications• Does not provide formal enough descriptive
representations for machine semantics– Methodology is useful there, too
15
Tekoäly, Eero Hyvönen, 2004
29
”Classical” thesauri• Semantically arranged
terminologies/dictionaries– Terms may be a mixture of words and concepts
• Based traditionally on the following relations– BT Broader term
NT Narrower termRT Related termUSE “See”UF Used for; reciproc. USESN Scope note
• E.g. in Finland:– Yleinen Suomalainen Asiasanasto, MASA, MUSA,
Allers, …
Tekoäly, Eero Hyvönen, 2004
30
• Used especially for– Indexing information content (keywords)– Information retrieval
• Keyword search• Term expansion: “tree” -> “pine”, “birch”,…
• Widely used, lots on indexed data– E.g. libraries, museums, archieves, …
16
Tekoäly, Eero Hyvönen, 2004
31
Limitations
• Meaning of relations– BT/NT for sub/superordinate, part-of etc.– RT has lots of different interpretations
• Cause/effect, tool/product, …• Cf. associative relation in terminology
• Not formal enough for computers– E.g. delimiting characteristics implicit– Semantics vague
Tekoäly, Eero Hyvönen, 2004
32
Esimerkki:Tesaurus vs. ontologia
– SisustusesineetST Peilit
PeilitST Taskupeilit
– Ok, mutta kyselyn ”Hae sisustusesineet”tulokseen tulevat myös taskupeilit!
17
Ontology
Computer Science Perspective
Tekoäly, Eero Hyvönen, 2004
34
• Information Systems– 1967 G. H. Mealy
• Relating data with the real world– Object-oriented programming
• The main paradigm in practice since the late 90’s
• Artificial Intelligence– Since 60’s– Natural Language Understanding – Knowledge Representation
• Logic + Ontologies + Computations
• WWW and the Semantic Web– Since late 90’s
18
Tekoäly, Eero Hyvönen, 2004
35
Ontologian käsite
• “Ontologia on formaali, eksplisiittinenmäärittely yhteisestä käsitteistöstä”(Gruber, 1993)
• Formaali: jämpti• Eksplisiittinen: jopa konekin voi ymmärtää• Yhteinen: kommunikaatio mahdollista
• Ontologia kuvaa– sovellusmaailmassa olevat käsitteet/oliot ja– niistä käytettävän sanaston
Tekoäly, Eero Hyvönen, 2004
36
• Komponentit– Käsitteiden määrittely: koneen ymmärtämällä
tavalla– Terminologia: ihmisten ymmärtämällä tavalla
• Edellytys sille, että ihmiset ja koneet voivatymmärtää toisiaan
19
Tekoäly, Eero Hyvönen, 2004
37
Glossary- word list- little structure
Thesaurus- relations- NT, BT, RT etc.
Taxonomy- relations- inheritance- constrains
Axiomatized theory- formal system- logic-based
Ontological complexity/depth
Hum
anun
ders
tand
able
Mac
hine
un
ders
tand
able ONTOLOGY TYPES
Numbers
Philosophical text
Tekoäly, Eero Hyvönen, 2004
38
MAO
abstraktit käsitteettoimijat
tapahtumatmateriaalit ja aineet
esineetarkisto- ja kirjastoaineisto
organismitympäristöt
AAT AAT ArtArt & & ArchitectureArchitecture ThesaurusThesaurus-- Paul Getty Paul Getty --ssäääätitiöön n thesaurusthesaurus-- 7 p7 pääääluokkaa, 125 000 kluokkaa, 125 000 kääsitettsitettää
20
Tekoäly, Eero Hyvönen, 2004
39
Museum ontology MAO
Tekoäly, Eero Hyvönen, 2004
40
IEEE Standard Upper MergedOntology (SUMO)
• Goals– Automated reasoning support in knowledge-based applications– Interoperability
• Define new data elements using SUMO and obtain mutualinteroperability
• Interoperability between applications using domain specificontologies (that use SUMO)
• Neutral interchange format for different systems
• Application areas– E-commerce– E-learning– Natural langugage understanding tasks– …
21
Tekoäly, Eero Hyvönen, 2004
41
SUMO Principal Distinctions
Tekoäly, Eero Hyvönen, 2004
42
SUMO Object:
22
Tekoäly, Eero Hyvönen, 2004
43
Standard Upper Merged Ontology
Tekoäly, Eero Hyvönen, 2004
44
SUMO Knowledge Representation• Developed in KIF (Knowledge Interchange
Format)– A version of first order predicate logic– Other versions exist (e.g. OWL)
• Size– 1006 terms– 4142 axioms– 814 rules
23
Tekoäly, Eero Hyvönen, 2004
45
Cyc Ontology• Idea: enCYClopeadic knowledge base
– Common sense knowledge• Goal: emergent intelligence by quantity (rather
than by quality)– Over 100.000 concepts– About 1.000.000 facts & axioms in logic
• www.cyc.com– Top parts of the ontology are available for free
• Garbage in -- garbage out?
Tekoäly, Eero Hyvönen, 2004
46
Cyc ontology
24
Tekoäly, Eero Hyvönen, 2004
47
Some Difficulties in Ontology Engineering
• How to select individuals, subclasses & properties?– GeorgeBush < … < Process
• How to define them?– Collection: Intangible, Imperciple?
• Scaling problems: 1,000,000 facts• Different needs in different applications
– How to reuse and share knowledge?
Tekoäly, Eero Hyvönen, 2004
48
• How to align or merge ontologies?• How to deal with changing ontologies?
– Versioning– Ontology time series
• How to deal with imprecise ontologies?
25
Tekoäly, Eero Hyvönen, 2004
49
Ontology reuse & sharing
– Special ontologies for specific microworld applications– E.g. River < Line
• but for a ship also river depth matters!– How share ontologies in different applications?
• More generic KB framework is needed
(Warren, Pereira, 1982)
Tekoäly, Eero Hyvönen, 2004
50
Ontology alignment
Know
Wissen = Savoir(knowing-that)
Kennen = Connaitre
River(size = big)
Fleuve(river running into sea)
Riviere(river or stream running into anotherriver)
Stream(size = small)
BigRiviere
(Sowa, 2001)
26
Tekoäly, Eero Hyvönen, 2004
51
Ontology alignment
(Sowa, 2004)
Tekoäly, Eero Hyvönen, 2004
52
Adjective-noun concepts– Prehending entity
• Property modifies directly the object (class instance)• “old man” = the person is old
– Prehended entity• Property modifies the class• “nuclear physicist” = the person isn’t “nuclear”
– Intention• modifies the relation• “former president” = the person is not former, but was the
president
27
Tekoäly, Eero Hyvönen, 2004
53
Example– “Sibelius is a good musician …
• good(Sibelius) & musician(Sibelius)– … and a bad cook”
• bad(Sibelius) & cook(Sibelius)– => Is Sibelius is a bad musician? -> Yes!
– The problem:• good/bad modify class musician/cook• musician/cook modify object Sibelius
Tekoäly, Eero Hyvönen, 2004
54
Ontology in computer science: what is needed?
• Formal expressive language for defining concepts and terms, their mutual relations, and inferences– Procedural systems: object-oriented programming
• Java, C++, …– Declarative systems (logic based)
• Logic: RDF(S), KIF, Prolog, OWL,…
• Ontological theory represented in it• E.g. Cyc
• (Application making use of (interpreting) the theory)
• Procedural semantics
1
3-Dec-04 1
Towards Intelligent Web Services
6
Bringing the web to its full potential
Semantic W
eb enabled W
eb Services
Static
Dynamic UDDI, WSDL, SOAPWeb Services
URI, HTML, HTTP RDF, RDF(S), OWLWWW Semantic Web
Intelligent WebServices
The General Vision
(Dieter Fensel, 2002)
3-Dec-04 2
Web Generations
1G WWW:WWW-pages for human interpretationHTML
2G WWW: Structures for human/machine interpretationXML
3G WWW: Semantic WebMeanings for human/machine interpretationRDF(S), OWL
=> New basis for intelligent WWW-services
2
3-Dec-04 3
Semantic Web & Web Services:
Teknologioita
3-Dec-04 4
Semantic Web:Technology push
Internet levelUnicode, URI,...
Structure levelXML, XML DTD/ Schema, XSL,...
Metadata levelRDF, Topic Maps,...
Ontology levelRDFS, OWL, WordNet, ...
Logic levelKIF, RuleML, SWRL, ...
Trust levelDigital signature, annotations,... Planning
CPR, SPAR, PDDL,…
ProcessesBPML, WPDL, PSL,…
ServicesUDDI, WSDL, OWL-S,…
TransactionsXML/EDI, KQML,…
CommunicationTCP/IP, HTTP, SOAP,...
Laajennettu Tim Berners-Leen (W3C) ”teknologiakakku”
3
3-Dec-04 5
Metadata level
3-Dec-04 6
RDF Resource Description FrameworkRDF draft 1999, recommendation 2.10.2004– W3C recommendation 10.2.2004– For representing metadata– Relational data model, not a syntax like XML
RDF-description = directed graphCan be serialized by XML
4
3-Dec-04 7
RDF(S)
RDF Resource Description Framework (1999)– General framework for describing web resources – Specification: Model & syntax– Relational model, not a syntax like XML
RDF Schema (2000)– For defining RDF- vocabularies– Object-oriented descriptions for WWW languages
Class hierarchies, properties, constraints, inheritance (Class/subClass/type)
3-Dec-04 8
RDF Model
Idea: metadata about WWW resources (URIs)Object-Attribute-Value triples– "Helsinki"-"is located in"-"Finland"– <http://www.helsinki.fi, Location, http://www.finland.fi>
Reification– “Hierarchical” statements about statements
5
3-Dec-04 9
RDF SyntaxXML-based (other serializations exist, too!)" The WWW-paper "Using RDF …" was written by John Smith, concerns RDF, Metadata and Semantic Web", and was created on Jan 1, 2000"
<?xml version="1.0"?><rdf:RDF xmlns:rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:DC = "http://purl.org/dc/elements/1.1/"> <rdf:Description rdf:about = "http://www.helsinki.fi/john.smith/paper.html"> <DC:Title>Using RDF to describe web resources</DC:Title> <DC:Creator>John Smith</DC:Creator> <DC:Date>2001-01-01</DC:Date> <DC:Subject>RDF, Metadata, Semantic Web</DC:Subject> </rdf:Description></rdf:RDF>
3-Dec-04 10
Corresponding RDF model (graph)
6
3-Dec-04 11
Creating RDF descriptions
RDF is a language for the machine, not for end-user (like HTML)RDF-generation:– From databases (hidden web)– From document metadata
E.g. PDF– From higher level languages
E.g. DAML+OIL, OWL ontology languages– By metadata editors
E.g. RDFPic
3-Dec-04 12
Why RDF is needed?
Generalizing the underlying model– XML model is a tree– RDF model is a graph
The arcs in graph do not have ordering (like in a tree)External metadata descriptions possibleGraphs can be merged
– Looks like a tiny detail but is an important stepGeneral "framework“, not domain specific
7
3-Dec-04 13
RDF Example
(Maedche, 2002)
3-Dec-04 14
RDF Schema
– W3C draft 2000, recommendation 10.2.2004– Defines RDF vocabularies– Object oriented modeling for WWW
Classes, instances, propertiesHierarchies, inheritance (Class/subClass/type)Property contraints (domain, range)
8
3-Dec-04 15
RDF(S) Example
(Maedche, 2002)
3-Dec-04 16
Ontology level
9
3-Dec-04 17
OWL Web Ontology LanguageW3C Standard– Next level above RDF(S)
Based on formal (description) logic– Inference, consistency– Subsumption: find objects satisfying a description– Subset of predicate logic
Optimized for subsumbtion relation & decidability
Human-friendly tools being developed– RDF(S) is produced by the machine
Based on– USA: DAML– EU: OIL
3-Dec-04 18
class-def animalclass-def plant
subclass-of NOT animalclass-def tree
subclass-of plantclass-def branch
slot-constraint is-part-of has-value treeclass-def leaf
slot-constraint is-part-of has-value branchclass-def defined carnivore
subclass-of animalslot-constraint eats value-type animal
class-def defined herbivoresubclass-of animalslot-constraint eats
value-type plant OR (slot-constraint is-part-of has-value plant)class-def herbivore
subclass-of NOT carnivoreclass-def giraffe
subclass-of animalslot-constraint eats
value-type leafclass-def lion
subclass-of animalslot-constraint eats value-type herbivore
class-def tasty-plantsubclass-of plantslot-constraint eaten-by has-value herbivore, carnivore
EXAMPLE OF AN OIL ONTOLOGY
10
3-Dec-04 19
Why OWL?Terminology logics for defining ontologiesUsage– Design phase
Check consistencyDerive subsumption hierarchy
– Data integrationDetect inconsistenties and unintendent relations
– DeploymentTerm expansion and inference, e.g. in information retrievalUsing descriptions in applications
Generic tools for cross-domain applications– E.g. Protégé OWL Plugin
Open standardW3C Recommendation
3-Dec-04 20
Three different versions of OWL– OWL Lite
Classifications + simple constraints– Local contraints on property values- cardinality 0 or 1
- OWL DL (description logic)- Complex classes
- unionOf, intersectionOf, compmentOf- Enumerated and disjoint classes
- oneOf, disjointWith- Classes defined on particular property values
- hasValue- OWL Full
- Classes as instances- Free cardinality
11
3-Dec-04 21
OWL example
<owl:Class rdf:ID="Wine"><rdfs:subClassOf rdf:resource=
"&food;PotableLiquid" />...<rdfs:subClassOf><owl:Restriction><owl:onProperty
rdf:resource="#hasMaker" /><owl:allValuesFrom
rdf:resource="#Winery" /></owl:Restriction>
</rdfs:subClassOf>...
</owl:Class>
Local property value contraints (not possible in RDFS)- Maker of Wine must be a Winery; maker of chairs smthg elseImporting food-ontology
3-Dec-04 22
OWL vs. RDFSClass definitions
Lattice hierarchy of concepts between classes Thing, NothingLogical combinations of anonymous classes
– unionOf, intersectionOf, complementOfEquality
Property restrictionsQuantification & class based -constraints
– allValuesFrom, someValuesFromCardinalityValue restrictions: hasValue
Property typesTransitive, Symmetric, Functional, Inverse, Equal
12
3-Dec-04 23
A Good OWL-tutorial and Tool
3-Dec-04 24
WHAT IS NEW?
Object-orientedmodeling
Knowledge representationLogic based semantics
XML-syntax, e.g., RDF(S)
PROGRAMMING ARTIFICIAL INTELLIGENCE
WWW-TECHNOLOGIES
13
3-Dec-04 25
Applications
Yhteentoimivuus (interoperability)Informaation haku (information retrieval)Tietämyksen hallinta (knowledge management)Web palvelut (Web Services) sovelluksineenProfilointi ja kustomointi (profiling, customizing)
3-Dec-04 26
Open Directory ProjectLinux-henkinen hanke tuottaa metadataa web-sisällöistä– yli 65.000 toimittajaa eri aloilla
Metatiedon yhdistely perustuu yhteiseen RDFS ontologiaan– 590.000 kategorian hierarkia!
Isot mittasuhteet – yli 4 miljoonaa annotoitua sivua
Yahoo-tyyppinen hakukone
(Tilanne 11.11.2004)
14
3-Dec-04 27
Open Directory Projecthttp://dmoz.org
3-Dec-04 28
15
3-Dec-04 29
Google ja ODP
3-Dec-04 30
16
3-Dec-04 31
TAP Ontology and News
3-Dec-04 32
TAP News and Related Stuff
17
3-Dec-04 33
Magpiehttp://kmi.open.ac.uk/projects/magpie/main.html
Ontologiat
3-Dec-04 34
Case: Promoottori– Photo repository of the Promotion Ceremonies
at the University of Helsinki– Exhibition to be opened at the Senate Square
at Helsinki city center 10/2003The Helsinki University Museum
– An interesting view into thelife of the university
– Complicated semantics of obscure events
– Novice users who do not know the contents before
– Team: Eero Hyvönen, Samppa Saarela, Kim Viljanen, Arvil Styrman, Jaana Tegelberg, …
1
12/3/2004
Case Case MuseoSuomiMuseoSuomihttphttp://://museosuomi.cs.helsinki.fimuseosuomi.cs.helsinki.fi
12/3/2004 © Eero Hyvönen
The Vision of The Vision of MuseumFinlandMuseumFinland
1.1. Global View to Distributed CollectionsGlobal View to Distributed Collections•• OneOne seamlessseamless national national collectioncollection ((virtuallyvirtually))•• ””MuseumsMuseums in Finland” in Finland” --> ”> ”MuseumMuseum of Finland”of Finland”
2.2. IntelligentIntelligent ServicesServices to to EndEnd--UsersUsers•• Search: ConceptSearch: Concept--Based Information RetrievalBased Information Retrieval•• Browsing: Semantically Linked ContentsBrowsing: Semantically Linked Contents
3.3. Easy Content Publication for MuseumsEasy Content Publication for Museums
CreatingCreating a national a national platformplatform and a and a processprocess for the for the museumsmuseums to to publishpublish theirtheir contentcontent togethertogether on on
the the SemanticSemantic WebWeb
2
12/3/2004 © Eero Hyvönen
1. Global View to Distributed 1. Global View to Distributed CollectionsCollections
Single accessSingle access--point to all collectionspoint to all collectionsSingle user interface to be learnedSingle user interface to be learnedTypicalTypical approachapproach todaytoday: : MultiMulti--searchsearch ((MetaMeta--searchsearch))–– E.gE.g. Australian . Australian MuseumsMuseums onlineonline
12/3/2004 © Eero Hyvönen
Australian Museums onlineAustralian Museums online
3
12/3/2004 © Eero Hyvönen
SearchSearch ResultResult
12/3/2004 © Eero Hyvönen
ProblemsProblems
HowHow to to findfind the the rightright keywordskeywords??LikelyLikely resultsresults: : nono--hitshits oror 1000 1000 hitshitsTooToo manymany irrelevantirrelevant hitshits ((lowlow precisionprecision))TooToo fewfew relevantrelevant hitshits ((lowlow recallrecall))HowHow to to getget overviewsoverviews of the of the contentscontents??HowHow to to findfind relatedrelated objectsobjects??Is Is thisthis entertainingentertaining & & educatingeducating??
4
12/3/2004 © Eero Hyvönen
MuseumFinlandMuseumFinland ApproachApproach
12/3/2004 © Eero Hyvönen
2. 2. IntelligentIntelligent ServicesServices to to EndEnd--UsersUsers
MuseumFinlandMuseumFinlandDemonstrationDemonstration
1.1. MultiMulti--facetfacet searchsearchBasedBased on on ontologiesontologies
2.2. SemanticSemantic browsingbrowsingBasedBased on on logiclogic
3.3. KeywordKeyword searchsearchBasedBased on on conceptsconcepts http://museosuomi.cs.helsinki.fi
5
12/3/2004 © Eero Hyvönen
MuseumFinlandMuseumFinlandMobile VersionMobile Version
WAP 2.0 WAP 2.0 compatiblecompatible phonesphonesNokia Nokia SeriesSeries 60 60 browserbrowserCombiningCombining coordinatescoordinatesand and locationlocation ontologyontologyGettingGetting moremore info info bybyemailemailSameSame URL:URL:http://http://museosuomi.cs.helsinki.fimuseosuomi.cs.helsinki.fi
12/3/2004 © Eero Hyvönen
3. Easy Local Content 3. Easy Local Content PublicationPublication
–– The simple publishing idea of the WWWThe simple publishing idea of the WWW–– The museum creates RDF and publishes it in a The museum creates RDF and publishes it in a
public directory (or on CD)public directory (or on CD)ContentContent creationcreation toolstools: : TerminatorTerminator, , AnnomobileAnnomobileContentContent creationcreation hashas itsits ownown challengeschallenges
6
12/3/2004 © Eero Hyvönen
Summary:Summary:Why Why MuseumFinlandMuseumFinland??MuseumMuseum Visitor’sVisitor’s viewpointviewpoint
SeamlessSeamless viewview to to heterogeneousheterogeneous collectionscollectionsIntelligentIntelligent ServicesServices
SearchSearch based on views and based on views and ontologiesontologiesBrowsingBrowsing based on semantic recommendationsbased on semantic recommendations
Museum’sMuseum’s viewpointviewpointPublicationPublication channelchannel for the for the SemanticSemantic WebWebOnlyOnly contentcontent needsneeds to to bebe createdcreated
WillWill bebe enhancedenhanced withwith new new ontologiesontologies and and materialsmaterials
MuseumFinlandMuseumFinland = = DemonstrationDemonstration on on howhow to to createcreate a national a national culture culture contentcontent publicationpublication channelchannel for the for the SemanticSemantic WebWeb
12/3/2004 © Eero Hyvönen
AwardsAwards to to MuseumFinlandMuseumFinland 20042004
7
12/3/2004 © Eero Hyvönen
FromFrom MuseumFinlandMuseumFinland to to CultureFinlandCultureFinland and and beyoundbeyound
New New kindskinds of of semanticsemantic materialmaterial–– ContentContent annotationannotation of of artart and and photosphotos
CurrentlyCurrently paintingspaintings onlyonly as as artifactsartifactsThe National The National GalleryGallery and ICONCLASSand ICONCLASSThe The FinnishFinnish MuseumMuseum of of PhotographyPhotography
–– ContentContent annotationannotation of (of (educationaleducational) ) videosvideosThe The FinnishFinnish BroadcasringBroadcasring Company (YLE)Company (YLE)
–– ContentContent annotationannotation of of narrativesnarratives and and processesprocessesThe national The national eposepos ”Kalevala””Kalevala”
Motto: ”Motto: ”MuseumFinlandMuseumFinland --> > CultureSampoCultureSampo””
12/3/2004 © Eero Hyvönen
ThanksThanks