CS652 Spring 2004 Summary

1
CS652 Spring 2004Summary
2
Course Objectives

• Learn how to extract, structure, and integrate
  Web information
• Learn what the Semantic Web is
• Learn how to build ontologies for the Semantic
  Web
• Investigate class-related research topics
• Be introduced to Semantic Web services

3
Generally Applicable Ideas

• Semantic Understanding
• Data attribute-value pairs
• Information data in a conceptual model
• Knowledge information with agreement
• Meaning useful knowledge
• Measuring Success
• Recall NrCorrect/TotalCorrect
• Precision NrCorrect/(NrCorrectNrIncorrect)
• F-measure (ß21)PR/(ß2PR)

4
Information Extraction

• Get relevant information
• Not
• Information retrieval get relevant pages
• Web mining discover unknown associations
• Wrapper maps data to a suitable format
• Generation techniques
• Machine learning (e.g. RAPIER)
• Natural language processing (e.g. RAPIER)
• Hidden Markov Models
• By-example generation tools (e.g. Lixto)
• By-pattern generation (e.g. RoadRunner)
• Wrapper Maintenance

5
Information Extraction BYU Ontos

• Ontology-based
• Data frames
• Strengths
• Resilient to page changes
• Robust across sites within the same domain
• Works well with all types of data-rich text
• Weaknesses
• Hand-crafted ontologies and data frames
• Requires record-boundary recognition
• Does not learn
• Applications
• Extraction
• High-precision classification
• Schema mapping
• Semantic Web annotation
• Agent communication
• Ontology generation

6
Semantic Web

• Tim Berners-Lee
• information has a well-defined meaning
• enables computers and people to work in
  cooperation
• Adds context and structure via metadata
• Agent computing paradigm
• Knowledge markup semantic annotation

7
Ontologies

• a formal, explicit specification of a shared
  conceptualization Gruber93
• Formal machine readable FOL
• Explicit concepts and constraints explicitly
  defined
• Shared community accepted
• Conceptualization abstract model (OSM)
• shared vocabulary

8
Ontology Formalism
Ontology O ltV, Agt where V vocabulary
predicate symbols (each with some arity) A
axioms formulas (constraints and rules)
Predicates Owner(x), Vehicle(x), Car(x),
Truck(x), Owner(x) owns Vehicle(y) Formulas
?x(Car(x)?Truck(x) ? Vehicle(x)) ?x(Owner(x) ?
??1y(Owner(x) owns Vehicle(y)) Inference Rules
TruckOwner(x) - Owner(x), Owner(x) owns
Vehicle(y), Truck(y)
9
Semantic Web Ontologies

• RDF
• DAMLOIL
• OWL

10
Semantic Web Annotationwith BYU Ontos
BYU Ontos Extraction Ontology
OWL Ontology
osm.cs.byu.edu/CS652s04/ontologies/OWL/carads.owl
Annotated Semantic Web Page
osm.cs.byu.edu/CS652s04/ontologies/annotatedPages/
carSrch1_semweb.html
11
Ontology Generation for the Semantic Web

• Necessary for the Semantic Web
• Ontology engineering
• Tools
• Methodology
• Languages (e.g. SHOE, OWL)
• Semiautomatic generation
• NLP machine learning (e.g. OntoText)
• Create from dictionary or lexicon (e.g. Doddle)
• Generation from tables (e.g. TANGO)
• Ontology maintenance

12
Ontology Libraries for theSemantic Web

• Locating ontologies
• Indexing and organization
• Search mechanisms
• Reusing ontologies
• Find one and modify
• Find several, merge and modify

13
Ontology Mapping, Merging, and Integration for
the Semantic Web

• Ontology reuse
• Heterogeneous agent communication
• Agent commitment to a new ontology
• On the fly map, merge, integrate (nontrivial to
  automate)
• Can we do well enough?
• Can we synergistically involve a user?
• Information extraction wrt target
• Table extraction (BYU Ontos)
• Semiautomatic wrapper/mediator construction by
  automatically providing mappings

14
Schema Mapping

• Schema-level matchers
• Name matchers (dictionaries WordNet)
• Structural context matchers
• Instance-level matchers
• Value characteristics
• Data-frame matchers
• Mapping cardinality
• 11 (direct)
• 1n, n1, nm (indirect, complex)
• Multi-faceted mapping techniques

15
Schema Integration

• FCA merge using lattices
• Global as View (GAV)
• Global mediator relations are views over source
  relations
• Dynamic mediator schema changes to accommodate
  new sources (hard to add new sources)
• Query only requires view unfolding
• Good for static, centralized systems
• TSIMMIS
• Local as View (LAV)
• Local source relations are views over mediator
  relations
• Fixed mediator schema new sources identify
  components covered (easy to add new sources)
• Complex query rewriting
• Good for dynamic, distributed systems
• Information Manifold

16
What is your dream for the Semantic Web?

• Intelligent personal agents that can
• Gather (just) the information we want and deliver
  it to us when we want it
• Help us with scheduling
• Help us buy the goods we want
• Negotiate and conduct business for us
• Intelligent business agents
• Intelligent discovery agents

What can you do to make your dreams come true?