Similitude entre les sens d’usage d’un terme dans un réseau lexical [Similarity between term senses in a lexical network]

1685

Authors

nan

Abstract

: Typed lexical relations between terms are indispensable for the tasks realized in NLP, but collecting lexical information is a difficult process. Indeed, when done manually, it requires the competence of experts and the duration can be prohibitive. When done automatically, the results can be biased by the chosen corpus of texts. The approach we present here consists in having people take part in a collective project by offering them a playful application accessible on the web. From an already existing base of terms, the players themselves thus build the lexical network, by supplying associations which are validated only by an agreeing pair of users. Furthermore, these typed relations are weighted according to the number of pairs of users who provide them. We then approach the question of word usage determination for a term, by searching relations between this term and its neighbours in the network, before introducing the notion of similarity between these different word usages. We are thus able to build the tree of word usages for a term. Finally, we briefly present the realization and the first obtained results. MOTS-CLÉS : traitement automatique du langage naturel, réseau lexical, relations typées pondérées, sens d’usage d’un terme, jeu en ligne. KEYWORDS : Natural Language Processing, lexical network, typed and weighted relations, word usage, web-based game. TAL. Volume 50 – n° 1/2009, pages 177 à 200 178 TAL. Volume 50 – n° 1/2009 1.

Keywords

nan

Full Content

Similitude entre les sens d’usage d’un terme dans un réseau lexical Mathieu Lafourcade – Alain Joubert LIRMM – Université Montpellier 2 - CNRS Laboratoire d’Informatique, de Robotique et de Microélectronique de Montpellier 161, rue Ada, F-34392 Montpellier cedex 5 {lafourcade, joubert}@lirmm.fr RÉSUMÉ : Les relations lexicales typées entre termes sont indispensables pour les tâches réalisées en TALN, mais leur collecte peut s’avérer difficile. En effet, effectuée manuellement, cette tâche nécessite la compétence d’experts et la durée nécessaire peut être prohibitive, alors que, réalisée automatiquement, les résultats peuvent être biaisés par les corpus de textes retenus. L’approche présentée ici consiste à faire participer un grand nombre de personnes à un projet contributif en leur proposant une application ludique accessible sur le Web. À partir d’une base de termes préexistante, ce sont ainsi les joueurs qui vont construire le réseau lexical, en fournissant des associations qui ne sont validées que si elles sont proposées par au moins une paire d’utilisateurs. De plus, ces relations typées sont pondérées en fonction du nombre de paires d’utilisateurs qui les ont proposées. Nous abordons ensuite la question de la détermination des différents sens d’usage d’un terme, en analysant les relations entre ce terme et ses voisins immédiats dans le réseau lexical, avant d’introduire la notion de similitude entre ces différents sens d’usage. Nous pouvons ainsi construire pour un terme son arbre des usages. Enfin, nous présentons brièvement la réalisation et les premiers résultats obtenus. ABSTRACT : Typed lexical relations between terms are indispensable for the tasks realized in NLP, but collecting lexical information is a difficult process. Indeed, when done manually, it requires the competence of experts and the duration can be prohibitive. When done automatically, the results can be biased by the chosen corpus of texts. The approach we present here consists in having people take part in a collective project by offering them a playful application accessible on the web. From an already existing base of terms, the players themselves thus build the lexical network, by supplying associations which are validated only by an agreeing pair of users. Furthermore, these typed relations are weighted according to the number of pairs of users who provide them. We then approach the question of word usage determination for a term, by searching relations between this term and its neighbours in the network, before introducing the notion of similarity between these different word usages. We are thus able to build the tree of word usages for a term. Finally, we briefly present the realization and the first obtained results. MOTS-CLÉS : traitement automatique du langage naturel, réseau lexical, relations typées pondérées, sens d’usage d’un terme, jeu en ligne. KEYWORDS : Natural Language Processing, lexical network, typed and weighted relations, word usage, web-based game. TAL. Volume 50 – n° 1/2009, pages 177 à 200 178 TAL. Volume 50 – n° 1/2009 1. Introduction Un très grand nombre d’applications en traitement automatique des langues (TAL), en particulier celles requérant un processus de désambiguïsation, nécessite la connaissance de relations lexicales ou …

Content truncated to 500 words

Labeled Text

: Typed lexical relations between terms are indispensable for the tasks realized in NLP, but collecting lexical information is a difficult process. Indeed, when done manually, it requires the competence of experts and the duration can be prohibitive. When done automatically, the results can be biased by the chosen corpus of texts. The approach we present here consists in having people take part in a collective project by offering them a playful application accessible on the web. From an already existing base of terms, the players themselves thus build the lexical network, by supplying associations which are validated only by an agreeing pair of users. Furthermore, these typed relations are weighted according to the number of pairs of users who provide them. We then approach the question of word usage determination for a term, by searching relations between this term and its neighbours in the network, before introducing the notion of similarity between these different word usages. We are thus able to build the tree of word usages for a term. Finally, we briefly present the realization and the first obtained results. MOTS-CLÉS : traitement automatique du langage naturel, réseau lexical, relations typées pondérées, sens d’usage d’un terme, jeu en ligne. KEYWORDS : Natural Language Processing, lexical network, typed and weighted relations, word usage, web-based game. TAL. Volume 50 – n° 1/2009, pages 177 à 200 178 TAL. Volume 50 – n° 1/2009 1.

Publication Details

Paper ID:
1685
Year:
2009
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan