McGuinnessAAAI July 28, 2002 The Semantic Web (State of the art and implications for language processing) Deborah McGuinness Associate Director and Senior.

McGuinness AAAI July 28, 2002

The Semantic Web (State of the art and implications for language

processing)

Deborah McGuinness

Associate Director and Senior Research Scientist

Knowledge Systems Laboratory

Stanford University

Stanford, CA USA

[email protected]

http://www.ksl.stanford.edu/people/dlm

McGuinness AAAI July 28, 2002

Outline

• Vision of the web today and tomorrow • The key to tomorrow’s web is semantics• Semantics on the web requires:

– Language for encoding meaning (DAML+OIL, OWL)

– Ontologies

– Tools

• Conclusion and Pointers

McGuinness AAAI July 28, 2002

Today: Rich Information Source for Human Manipulation/Interpretation

Human

Human

McGuinness AAAI July 28, 2002

“I know what was input”

• Global documents and terms indexed and available for search• Search engine interfaces• Entire documents retrieved according to relevance (instead of

answers)• Human input, review, assimilation, integration, action, etc.• Special purpose interfaces required for user friendly applications

The web knows what was input but does little interpretation, manipulation, integration, and action.

Analogous to a new assistant who is thorough yet lacks common sense, context, and adaptability

McGuinness AAAI July 28, 2002

Tomorrow: Rich Information Source for Agent Manipulation/Interpretation

Human

Agent

Agent

McGuinness AAAI July 28, 2002

“I know what was meant”

• Understand term meaning and user background• Interoperable (can translate between applications)• Programmable (thus agent operational)• Explainable (thus maintains context and can adapt)• Capable of filtering (thus limiting display and

human intervention requirements)• Capable of executing services

McGuinness AAAI July 28, 2002

Layer Cake Foundation

McGuinness AAAI July 28, 2002

Stated goals of Semantic Web

• Define conventions for applications that exchange metadata on the Web

• Enable vocabulary semantics to be defined by communities of expertise, not W3C

• Provide for the fine-grained mixing of diverse metadata

• Making it cost-effective for people to effectively record their knowledge.

• Ultimate goal - the design of enabling technologies to support machine facilitated global knowledge exchange

McGuinness AAAI July 28, 2002

Semantic Markup

Languages such as OWL, DAML+OIL(http://www.w3.org/2001/sw/WebOnt/, http://www.daml.org)• Encoding background info• User modeling info• Annotating web pages• Annotating services thereby limiting needs for human disambiguation input, human

interpretation, multiple answer display, translation assistance, agent assistance, adaptivity support, etc.)

Ontologies

DAML/OWL-enabled web

pages

McGuinness AAAI July 28, 2002

XML• World Wide Web Consortium (W3C)

standard• Provides important solution to syntax

problem and simple semantics and schemas:

<SSN>555-17-1234</SSN>

• Now we can describe the meaning of words• Many applications of XML appearing:

– Geographic Markup Language (GML)– Extensible rights Markup Language (XrML)– Chemical Markup Language (CML)

Problem: Limited semantics, limited ontology creation

McGuinness AAAI July 28, 2002

DARPA Agent Markup Language

• http://www.daml.org/about.html

• Extends the vocabulary of XML and RDF/S

• Provides rich ontology representation language

• Language features chosen so language may have efficient implementations

McGuinness AAAI July 28, 2002

DARPA DAML Program

• Began in August 2000 Kickoff meeting• 19 Research groups supported• Initial ontology language aims to extend

XML, RDF/S, benefit from frames, benefit from principled KR systems like Description Logics

• DAML-ONT released in Oct. 2000• DAML+OIL released in March 2001

McGuinness AAAI July 28, 2002

DAML Language

Web LanguagesRDF/SXML

DAML-ONT

Formal FoundationsDescription Logics

FACT, CLASSIC, DLP, …

Frame Systems

DAML+OIL(OWL)

OIL

McGuinness AAAI July 28, 2002

DAML+OIL -> W3C

• W3C Webont working group formed with DAML+OIL submission as starting point http://www.w3.org/Submission/2001/12/

McGuinness AAAI July 28, 2002

W3C Web Ontology Working Group

• Web Ontology Working Group in the W3C Semantic Web Activity aimed at “extending the semantic reach of current XML and RDF meta-data efforts. “

• History– W3C Announcement in November 2001 -

http://lists.w3.org/Archives/Public/www-rdf-logic/2001Nov/0000.html

– Weekly teleconferences starting in November 2001

McGuinness AAAI July 28, 2002

WEBONT catches on….

• Includes over 50 members from over 30 organizations. – Industry including:

• Large companies such as Daimler Chrysler, EDS, Fujitsu, HP, IBM, Intel, Lucent, Nokia, Philips Electronics, Sun, Unisys, …

• Newer/smaller companies such as IVIS Group, Network Inference, Stilo Technology, Unicorn Solutions, …

– Government and Not-For-Profits:• Defense Information Systems Agency, Interoperability Technology

Association for Information Processing, Japan (INTAP) , Intelink Mgt Office, Mitre, …

– Universities and Research Centers:• University of Bristol, University of Maryland, University of Southamptom,

Stanford University, …• DFKI (German Research Center for Artificial Intelligence),

Forschungszentrum Informatik– Invited Experts

• Well-known academics from non-W3C members

McGuinness AAAI July 28, 2002

WEBONT cont.

• Quarterly Face to Face meetings in – Murray Hill: http://www.w3.org/2001/sw/WebOnt/ftf1.html

– Amsterdam: http://www.w3.org/2001/sw/WebOnt/ftf2.html

– Stanford: http://www.w3.org/2001/sw/WebOnt/ftf3.html

– Upcoming – Bristol, and xxx

• Interesting Documents:– DAML+OIL submission – full spec with reference description,

walkthrough, FOL and model theoretic semantics, http://www.w3.org/TR/daml+oil-reference

– Use Case and requirements document: http://www.w3.org/TR/webont-req/

McGuinness AAAI July 28, 2002

OWL Lite and OWL

• Feature Synopsis:

http://www.w3.org/TR/owl-features/• Reference Description:

http://www.w3.org/TR/owl-ref/

• Abstract Syntax:

http://www.w3.org/TR/owl-absyn/

McGuinness AAAI July 28, 2002

OWL Lite Features•

RDF Schema Features– Class – rdf:Property – rdfs:subClassOf – rdfs:subPropertyOf – rdfs:domain – rdfs:range – Individual

• Equality and Inequality– sameClassAs – samePropertyAs – sameIndividualAs – differentIndividualFrom

• Restricted Cardinality – minCardinality (restricted to 0 or 1) – maxCardinality (restricted to 0 or 1) – cardinality (restricted to 0 or 1)

McGuinness AAAI July 28, 2002

OWL Lite Features (cont)• Property Characteristics

– inverseOf – TransitiveProperty – SymmetricProperty – FunctionalProperty(unique) – InverseFunctionalProperty (unambiguous) – allValuesFrom (universal local range restrictions; previously toClass) – someValuesFrom (existential local range restrictions; previously

hasClass)• Datatypes

– Following the decisions of RDF Core. • Header Information

– imports – Dublin Core Metadata – versionInfo

McGuinness AAAI July 28, 2002

OWL Features

• Class Axioms– oneOf (enumerated classes) – disjointWith – sameClassAs applied to class expressions – rdfs:subClassOf applied to class expressions

• Boolean Combinations of Class Expressions – unionOf – intersectionOf – complementOf

• Arbitrary Cardinality – minCardinality – maxCardinality – cardinality

• Filler Information– hasValue Descriptions can include specific value information

McGuinness AAAI July 28, 2002

Tools

In order to take off, we need tools: http://www.daml.org/tools/

Annotation Ontology TranslationBrowser PersistenceCrawler Query ToolsEditor RDMS MappingGraph Visualizer Report GenerationTransformation Search Validator Ontology AnalyzerImporter Ontology EditorInference Engine

Many are in research labs, but some in companies…Network Inference, Sandpiper, Ontoprise, ….

McGuinness AAAI July 28, 2002

Ontologies Exploding

• Upper Level Ontologies Emerging: – UNSPSC, SUO, OpenCyc, OpenDirectory,

TAP, …

• Specialized Ontologies Emerging– UMLS, NPC,

• Libraries– http://www.daml.org/ontologies/– http://www.ksl.stanford.edu/ontolingua

McGuinness AAAI July 28, 2002

Some Observations…

• Markup Languages are growing in acceptance and expressive power

• User base, tool base, ontology base growing

• Ontology-enhanced applications springing up (not just in ivory towers like FindUR, eCyc, …)

McGuinness AAAI July 28, 2002

Simple Ontology-Enhanced Apps

McGuinness AAAI July 28, 2002

McGuinness AAAI July 28, 2002

Applied SemanticsApplied Semantics uses a large scale ontology, or knowledge base

of concepts and their relationships, to bring semantic understanding

to the processing of unstructured information. Our software products

and services improve the business processes for publishing,

enterprise applications, and internet infrastructure markets by

automating content tagging, categorization, and summarization for

more effective information sharing and retrieval.

• Founded in 1998

• Won Internet World Fall 1999 “Best of Show” for meaning-based search

• 40 employees

• Funding from: Zero Gravity, Ridgestone, others

• 50+ customers, including:

McGuinness AAAI July 28, 2002

CIRCA in Publishing

USERS

USERS

News Content

CIRCA Technology

Auto-Categorizer

Metadata Creator

Page Summarizer

Proprietary Content

Management System

A major newspaper uses Auto-Categorizer by IPTC code (standard publishing taxonomy), Metadata Creator to generate meaningful thematic keywords, and Page Summarizer to summaries of varying lengths.

McGuinness AAAI July 28, 2002

Conclusion/Discussion

• The Semantic Web is in its infancy today but is ready for applications

• Markup Language, Ontologies, and some tools are ready for use

• Hybrid applications may be the first to grow like ontology-enhanced search, ontology-enhanced knowledge capture, etc.

• Lets get together…..

McGuinness AAAI July 28, 2002

Extras

McGuinness AAAI July 28, 2002

What is an Ontology?

Catalog/ID

GeneralLogical

constraints

Terms/glossary

Thesauri“narrower

term”relation

Formalis-a

Frames(properties)

Informalis-a

Formalinstance

Value Restrs.

Disjointness, Inverse, part-

of…

McGuinness AAAI July 28, 2002

Some Pointers

• Ontologies Come of Age Paper: http://www.ksl.stanford.edu/people/dlm/papers/ontologies-come-of-age-abstract.html

• OWL: http://www.w3.org/TR/owl-features/ , http://www.w3.org/TR/owl-ref/

• DAML+OIL: http://www.daml.org/ , http://www.w3.org/TR/daml+oil-reference

McGuinness AAAI July 28, 2002

Contact Information

[email protected]

www.ksl.stanford.edu/people/dlm