Sémantique interprétative et analyses automatiques de textes : que deviennent les sèmes ?

Abstract : Semes are usually understood as semantic primitives, in order to describe meanings. As for F. Rastier, semes are contextually defined, and are the result of an interpretative path. This modelisation accounts for successes and limits of some statistical and distributional approaches. Some of the mathematical properties implied are not linguistically appropriate, so that the processing has to be defined on new bases. Text semantics also sheds new light on NLP tools and processes : dictionaries, taggers, corpus. A new clustering method is proposed as a means to modelize isotopies, and thus texts topics. These computed semantic classes are inherently different from the ones manually defined by experts.
Liste complète des métadonnées

https://halshs.archives-ouvertes.fr/halshs-00367164
Contributor : Bénédicte Pincemin <>
Submitted on : Tuesday, March 10, 2009 - 3:07:24 PM
Last modification on : Friday, January 4, 2019 - 5:33:00 PM
Document(s) archivé(s) le : Thursday, June 30, 2011 - 11:07:54 AM

File

pincemin_semiotiques02.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : halshs-00367164, version 1

Collections

Citation

Bénédicte Pincemin. Sémantique interprétative et analyses automatiques de textes : que deviennent les sèmes ?. Sémiotiques, 1999, pp.71-120. 〈halshs-00367164〉

Share

Metrics

Record views

381

Files downloads

1068