SEMANTICS

Processing language at the semantic level means going beyond the particular words and syntactic constructions that happen to be used in a document, to identify the essential concepts being referred to and the relationships that are asserted among them. The abstraction provided by such representations allows an application to be more precise, for example distinguishing between uses of the word "bank" as a financial institution and as the edge of a river. Semantic representations also lend themselves to inference, particularly the use of background knowledge to refine or extend the interpretation of a text. For example, background knowledge about traffic accidents could be used to match a text about an accident to a query about people being injured, even if the text doesn't explicitly mention injuries.

ONGOING ACTIVITIES

Ontologies, knowledge representation, inference, contact: A. Kaplan

Discourse representation, contact: S. Aït-Mokhtar

Semantic normalisation, contact:  C. Brun and C. Hagège

Parsing and XML, C. Roux

Information extraction in the biological domain, contact: A. Kaplan and A. Sandor

PUBLICATIONS

• References to Semantics



• Publications by team members on semantics

PEOPLE

• Permanent: Frédérique Segond, Salah Ait-Mokhtar, Caroline Brun, Caroline Hagege, Aaron Kaplan, Claude Roux, Agnes Sandor.