Quand: le lundi 25 juin 2018, à 14h00
Où:Amphi. Etoile, à Télécom SudParis (Evry)

Titre: Natural Language Access to Knowledge: Where Reasoning Makes Sense

Richard Waldinger
Artificial Intelligence Center
SRI International
Menlo Park, California

Abstract:
Natural language is one of the more appealing ways by which people can interact with
computers, but up to now its application has been severely stilted. Natural language is
famously ambiguous; the same expression can have hundreds of possible meanings. On the other hand, there can be many ways of saying the same thing. Problems with natural language processing can be resolved by providing a system with an understanding of the subject domain. We approach these difficulties by equipping the system with the capability to perform deductive inference over a body of subject domain knowledge, expressed as an axiomatic theory. We illustrate this approach with SAP QUEST, a proof-of-concept system that allows naturallanguage question answering over a body of structured data for a business enterprise application. But we expect the approach to be more generally applicable. Joint work with Cleo Condoravdi, Kyle Richardson, Vishal Sikka, and Asuman Suenbuel

Bio:
Richard Waldinger is a Principal Scientist at SRI’s Artificial Intelligence Center, where he has worked since 1969. His interests focus on applying automatic deduction to problems in artificial intelligence and software engineering. He has worked in automatic program synthesis, planning, formal methods, and natural language understanding. He is the winner (with Zohar Manna) of the 2016 Herbrand Award, for Automatic Reasoning.