Disciplines/fields: Computational Linguistics, Computational Logic

Duration: 4 sessions

Course Content

  1. Logic as a model for the meaning of natural language; the method of fragments.
  2. World Knowledge and Inference (Tableaux and Model Generation)
  3. Extending Logic to deal with noun/verb phrases (Montague Grammar)
  4. Discourse Phenomena

Objectives

Current (statistical) models of natural language have trouble modeling the propositional level (truth/falsity/entailment) of meaning. Logical models provide this natively, but have scalability problems.

Conceptual: a basic understanding of logics and inference processes as a modelling paradigm with a propositional level.

Methodological: logics as specialized models, logic engineering, compositional natural language semantics construction, and inference in semantic-pragmatic analysis.

Literature

Blackburn/Bos: Representation and Inference for Natural Language: http://www.let.rug.nl/bos/comsem/book1.html

Vita

Dr. Michael Kohlhase is full professor for Computer Science at Jacobs University Bremen and an associate adjunct professor at Carnegie Mellon University.

He studied pure mathematics at the Universities of Tübingen and Bonn (1983 - 1989) and continued with computer science, in particular higher-order unification and automated theorem proving (Ph.D. 1994, Saarland University).

His current research interests include knowledge representation for mathematics, inference-based techniques for natural language processing, computer-supported education, and user assistance. He pursues these (interrelated) topics focusing on the aspects of modular foundations (usually logical methods) and large-scale structures in document corpora. He has pursued these interests during extended visits to Carnegie Mellon University, SRI International, and the Universities of Amsterdam, Edinburgh, and Auckland.

Link

http://kwarc.info/kohlhase