Titre :

Vers une analyse des différences interlinguistiques entre les genres textuels : étude de cas basée sur les n-grammes et l'analyse factorielle des correspondances

Auteur(s) :

Lefer, Marie-Aude [Auteur]
Université Catholique de Louvain = Catholic University of Louvain [UCL]
Bestgen, Yves [Auteur]
Université Catholique de Louvain = Catholic University of Louvain [UCL]
Grabar, Natalia [Auteur]
Savoirs, Textes, Langage (STL) - UMR 8163 [STL]

Titre de la manifestation scientifique :

TALN 2016: Traitement Automatique des Langues Naturelles

Ville :

Paris

Pays :

France

Date de début de la manifestation scientifique :

2016-07-04

Mot(s)-clé(s) en anglais :

Comparable Corpora
Correspondence Analysis
N-grams
Genres

Discipline(s) HAL :

Informatique [cs]/Traitement du texte et du document
Sciences de l'Homme et Société/Linguistique

Résumé en anglais : [en]

The aim of the present study is to assess the use of n-grams and Correspondence Analysis (CA) to compare genres in cross-linguistic studies. The study is based on an English-French bilingual corpus made up of original (i.e. ...
Lire la suite >The aim of the present study is to assess the use of n-grams and Correspondence Analysis (CA) to compare genres in cross-linguistic studies. The study is based on an English-French bilingual corpus made up of original (i.e. non-translated) texts, representing three genres: European parliamentary debates, newspaper editorials and academic articles. First, 2- to 4-grams are extracted in each language. Second, the most frequent 1000 n-grams for each n-gram length and in each language are analyzed by means of CA with a view to determining which n-grams are particularly salient in the genres examined. Finally, n-grams are manually classified into a range of categories, such as stance expressions, discourse markers and referential expressions. The results show that the n-gram approach makes it possible to uncover typical features of the three genres investigated, as well as interesting contrasts between English and French.Lire moins >

Langue :

Français

Comité de lecture :

Oui

Audience :

Internationale

Vulgarisation :

Non

Collections :

• Savoirs, Textes, Langage (STL) - UMR 8163

Source :

Harvested from HAL