Symbiotic Relationships between Semantic Web and knowledge engineering - PowerPoint PPT Presentation

1 / 9
About This Presentation
Title:

Symbiotic Relationships between Semantic Web and knowledge engineering

Description:

Symbiotic Relationships between Semantic Web and knowledge engineering ... Role of thesauri as light-weight ontologies as target ... Wiki's are not a panacea ... – PowerPoint PPT presentation

Number of Views:32
Avg rating:3.0/5.0
Slides: 10
Provided by: guus7
Category:

less

Transcript and Presenter's Notes

Title: Symbiotic Relationships between Semantic Web and knowledge engineering


1
Symbiotic Relationships between Semantic Web and
knowledge engineering
  • Impressions from Workshop 7

2
Administrative
  • Chairs Mark Musen Derek Sleeman
  • 13 papers
  • US/Canada 3
  • Europe 9
  • US/Europe collaboration 1
  • Format
  • Session of 90 minutes 2x 30 min papers, 30 min
    discussion
  • 1Invited speaker Alan Rector

3
Application dolmans
  • Majority
  • Medicine, biomedicine, microbiology
  • Others
  • Engineering, digital heritage
  • Application domain has strong influence on
    perspective on ontology engineering
  • e.g. on set of OWL constructs used
  • Role of thesauri as light-weight ontologies as
    target for enrichment and mapping

4
Collaborative knowledge engineering
  • Central theme, many projects tackle this (Halo,
    CRAFT, BioPortal, Collaborative Protégé, CleOn)
  • Bottom-up vs. top-down
  • Needs for annotations, work flows, differences of
    perspectives, mappings

5
Wikis and (collaborative) knowledge engineering
  • Wikis are not a panacea
  • Fulfill a limited but useful role, e.g. in
    collaborative taxonomy building
  • Web style of editing is beneficial for users
  • Special interfaces still used for rules an
    constraints
  • Hit them high and hit them low

6
Ontology quality
  • Ontology police?
  • Quality is not an objective characteristic
    depends on perspective and use
  • Good ontologies are used?! Or ontologies that
    are not used are bad ontologies
  • Ontology engineering sometimes requires a small
    team for boot strapping (instead of large
    collaboration)
  • Modularization important principle

7
Reasoning
  • Ontologies seem to be the fashion
    problem-solving methods ( reasoning patterns)
    are not en vogue (although there is ongoing work)
  • OWL DL classification reasoning used for ontology
    validation
  • Variety of general reasoning methods (e.g.
    adductive)

8
Applications
  • Examples
  • Information retrieval
  • biomedical information
  • multimedia collections
  • Proving hypothesis in functional genomics
  • Web-based knowledge-based systems

9
Ontology representation
  • OWL is de facto standard
  • OWL DL appears to be the focus in the biomedical
    area this is not the case in the other domains
  • Rules are used a lot SWRL is mentioned most
Write a Comment
User Comments (0)
About PowerShow.com