Pragmatic Semantics for the Web of Data

  • 322 views
Uploaded on

Presented at the AAAI Fall Symposium for Big Data on 2013-11-15.

Presented at the AAAI Fall Symposium for Big Data on 2013-11-15.

More in: Education , Technology
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
    Be the first to like this
No Downloads

Views

Total Views
322
On Slideshare
0
From Embeds
0
Number of Embeds
4

Actions

Shares
Downloads
1
Comments
0
Likes
0

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. PraSem A Pragmatic Semantics for the Web of Data Stefan Schlobach Wouter Beek (www.wouterbeek.com)
  • 2. Problem statement • The Web of Data (WoD) is complex, inherently messy, contextualised, and opinionated. • Today the WoD is constructed and used as a database. • Tomorrow the WoD should be constructed and used as a marketplace of ideas / a ‘knowledge economy’.
  • 3. Illustrative example
  • 4. Existing solutions for semantics • Context-dependence (contexts) • Complexity (small dataset + rich semantics, big datasets + less rich semantics) • Dynamicity (irregular snapshots) • In- or para-consistency (maximally consistent subset, reasoning light) • Objectivity (contexts, provenance) • Vagueness (Fuzzy logic)
  • 5. Alternative solution: Pragmatic Semantics Theory: • A collection of truth orderings, each representing a particular ‘worldview’. • A framework for optimisation over those truth-orderings. Implementation: • Distributed and nature-based algorithms.
  • 6. Examples of truth orderings • Model-theoretic notions of truth • (Classical) truth value • Ratio of maximally consistent subsets • Number of justifications • Structural aspects of the graph • Shortest path ordering (e.g. using random-walk distance) • Edge-weights • Node-ranks (e.g. PageRank) • Meta-data: • Popularity / abnormality / scarcity • Background knowledge from other sources: • Google count • Similarity / relevance
  • 7. Example At the VU university: • Computer Scientists talk about ‘ontologies’ • Philosophers talk about ‘ontology’ Suppose someone (foolishly?) asserted that a CS ontology is a Phil. ontology… • The deductive closure may contain falsities (e.g. “there is exactly one CS ontology”). But Computer Scientists are more connected with other Computer Science researchers than with Philosophers. When deduction is constrained by a structural metric, false assertions are less likely to arise.
  • 8. Pragmatic entailment
  • 9. Swarm intelligence
  • 10. Implementations Ant calculus: • Identify popular resources by random-walks, simulating PageRank. Bee calculus: • Dataset enrichment