NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé
For NLP systems, a major issue consists in resolving reference in order to find the themes of documents. In this article, we present a way to find semantic informations to resolve anaphors which use a referential expression of the form NdeN and an anaphor realized by a subject relative pronoun. This...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Cercle linguistique du Centre et de l'Ouest - CerLICO
|
Series: | Corela |
Subjects: | |
Online Access: | https://journals.openedition.org/corela/332 |
_version_ | 1797313856185303040 |
---|---|
author | Laurence Kister Evelyne Jacquey |
author_facet | Laurence Kister Evelyne Jacquey |
author_sort | Laurence Kister |
collection | DOAJ |
description | For NLP systems, a major issue consists in resolving reference in order to find the themes of documents. In this article, we present a way to find semantic informations to resolve anaphors which use a referential expression of the form NdeN and an anaphor realized by a subject relative pronoun. This method is based on the semantic content of nouns in NdeN groups and uses a list of concrete nouns in French which can be automatically extracted from the definitions of the TLFi dictionary in its XML-tagged version. Such a list is then used to annotate corpora in order to predict the selection of the good nominal referent which can be the second noun or the entire NP (35% of referential expressions contain a concrete noun). |
first_indexed | 2024-03-08T02:37:39Z |
format | Article |
id | doaj.art-7bd46f251ade4f29a1ee61b411e346ad |
institution | Directory Open Access Journal |
issn | 1638-573X |
language | English |
last_indexed | 2024-03-08T02:37:39Z |
publisher | Cercle linguistique du Centre et de l'Ouest - CerLICO |
record_format | Article |
series | Corela |
spelling | doaj.art-7bd46f251ade4f29a1ee61b411e346ad2024-02-13T13:52:41ZengCercle linguistique du Centre et de l'Ouest - CerLICOCorela1638-573X5110.4000/corela.332NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française InformatiséLaurence KisterEvelyne JacqueyFor NLP systems, a major issue consists in resolving reference in order to find the themes of documents. In this article, we present a way to find semantic informations to resolve anaphors which use a referential expression of the form NdeN and an anaphor realized by a subject relative pronoun. This method is based on the semantic content of nouns in NdeN groups and uses a list of concrete nouns in French which can be automatically extracted from the definitions of the TLFi dictionary in its XML-tagged version. Such a list is then used to annotate corpora in order to predict the selection of the good nominal referent which can be the second noun or the entire NP (35% of referential expressions contain a concrete noun).https://journals.openedition.org/corela/332lexical semanticsanaphorNdeNsemantic tagginglexicon extraction |
spellingShingle | Laurence Kister Evelyne Jacquey NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé Corela lexical semantics anaphor NdeN semantic tagging lexicon extraction |
title | NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé |
title_full | NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé |
title_fullStr | NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé |
title_full_unstemmed | NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé |
title_short | NdeN et acquisition d’informations lexicales à partir du Trésor de la Langue Française Informatisé |
title_sort | nden et acquisition d informations lexicales a partir du tresor de la langue francaise informatise |
topic | lexical semantics anaphor NdeN semantic tagging lexicon extraction |
url | https://journals.openedition.org/corela/332 |
work_keys_str_mv | AT laurencekister ndenetacquisitiondinformationslexicalesapartirdutresordelalanguefrancaiseinformatise AT evelynejacquey ndenetacquisitiondinformationslexicalesapartirdutresordelalanguefrancaiseinformatise |