Semantic Web and Agents

1. Semantic Web and Agents 許永真 國立台灣大學資訊工程學系

2. Information Revolution • Computers for scientific computation • Computers as productivity tools • Personal and home computers • Learning and research tools • Inter-person communication • Home entertainment • Information appliances • Explosive growth of the Web • 1000 users in 1990. • Over 1 billion users device-wide by 2003.

3. Common Uses of the Web • Information search and retrieval • News and entertainment • Shopping • Travel: reservation/purchase/check-in • Online community • B2B applications • P2P applications

4. A Futuristic Scenario • During your daily workout, you heard a beautiful song playing over the Internet radio. Unfortunately, your cell phone signaled that you had a call from an important client. Before picking up the call, you commanded your PDA immediately to get a copy of the song for your personal collection so that you can enjoy it at a later time. Meanwhile, the volume of the radio turned down automatically.

5. Behind the Scene • Your PDA will • Retrieve the current play list from the radio station. • Discover the song for purchase at several online stores. • Compare the price/feature and decide on a vendor. • Purchase the song on your behalf w/ your online wallet. • Schedule the song to be played later. • Meanwhile, • Your cell phone will announce your picking up the call. • Your radio will adjust the volume based on pre-defined conditions on its volume control.

6. What’s Missing? • Taskable agents • Goals, believes, & plans • Making decisions and taking actions for their users. • Machine-readable information sources • Mechanism for utilizing unstructured, heterogeneous, & distributed information • Communication among web-enabled devices • Identification of trusted services/information

7. Web  Semantic Web • Most of the Web’s content today is designed for humans to read. • Computers have no reliable way to manipulate & process the semantics. • The Semantic Web will bring structure to the meaningful content of web pages so software agents can carry out sophisticated tasks for users.

8. The Semantic Web “The Semantic Web is an extension of the current web in which information is given well-defined meaning, better enabling computers and people to work in cooperation.” • Tim Berners-Lee, James Hendler, Ora Lassila, The Semantic Web, Scientific American, May 2001

9. Evolution of Information Processing • Information becomes more • Complex • Structured • Meaningful • Processing becomes increasingly • Distributed • Interactive • Intelligent

10. Program Program Data Data Data Stand-Alone Processing • Program = Algorithm + Data • Information is stored in data files. • Programs may share information only by using data syntax defined for each data file.

11. Program Program DBMS DB DB DB Database • Information as structured data: records • Data schema & query language

12. KBMS KB: Facts + Rules Knowledge Base • Semantics: terms & relations • Rules of inference • Knowledge base query language

13. KR vs. Web • Limited expressiveness  anything goes! • Brittle and demand consistency  paradoxes and unanswered questions • Facts  non-sterilized information • Centralized  decentralized • Domain-specific  General • Small  scalable • Declarative data  Multimedia data

14. Wrapper Wrapper Wrapper IS IS IS Querying the Web • Search engines • Web directories • Information extraction w/o semantics • Wrappers Query Agent Mediator

15. Web  Knowledge Base • Goal: To develop a probabilistic, symbolic knowledge base that mirrors the content of the world wide web. If successful, this will make text information on the web available in computer-understandable form, enabling much more sophisticated information retrieval and problem solving. • Approach: developing a system that can be trained to extract symbolic knowledge from hypertext, using a variety of machine learning methods. • [Mitchell et al., 1998]

16. Semantic Web (Definition from W3C) • The Semantic Web is the abstract representation of data on the World Wide Web, based on the Resource Description Framework (RDF) standards and other standards to be defined. • It is being developed by the W3C, in collaboration with a large number of researchers and industrial partners.

17. Semantic Web “Layer Cake” [Tim Berners-Lee, XML 2000]

18. Semantic Web Agent • Web content that is meaningful to computers • Machine-readable ontology By Miguel Salmeron

19. RETSINA Calendar Agent To schedule meetings between individuals based on their schedules maintained in MS Outlook • Distributed Meeting Scheduling Engine • RETSINA Semantic Web Calendar Parser

20. Semantic Web Services

21. report.tar.gz Host A Host B pdf2ps uncompress gunzip uncompress latex gunzip dvips lpr pdf2ps Host C Host D

22. Service Agent Server Task Agent Interface Agent Task Agent Client Server Service Agent Task Agent Task Agent Server Service Agent Multi-Agent Architecture

23. XML Information Integration Agent [Jeong & Hsu, 2001]

24. Trust and Proofs [James Hendler, 2001]

25. Summary • The power of semantic web depends on successful integration of research on multiple disciplines: • Artificial intelligence • Information retrieval • Linguistics • Distributed systems & Web technology • Semantic web and agents will empower machines to collaborate with humans more effectively. • Agents will exploit users’ constraints and preferences to help customize users’ requests for automatic Web service discovery, execution, or composition and interoperation.