ScienQuest: a treebank exploitation tool for non NLP-specialists

Abstract : The exploitation of syntactically analysed corpora (or treebanks) by non NLP-specialist is not a trivial problem. If the NLP community wants to make publicly available corpora with complex annotations, it is imperative to develop simple interfaces capable of handling advanced queries. In this paper, we present query methods developed during the Scientext project and intended for the general public. Queries can be made using forms, lemmas, parts of speech, and syntactic relations within specific textual divisions, such as title, abstract, introduction, conclusion, etc. Three query modes are described: an assisted query mode in which the user selects the elements of the query, a semantic mode which includes local pre-established grammars using syntactic functions, and an advanced search mode where the user create custom grammars
Type de document :
Communication dans un congrès
Conférence COLING 2012, 2012, Mumbai, India. 2012
Liste complète des métadonnées

https://hal.inria.fr/hal-00954045
Contributeur : Gilles Sérasset <>
Soumis le : vendredi 28 février 2014 - 16:11:10
Dernière modification le : lundi 15 janvier 2018 - 21:46:02

Identifiants

  • HAL Id : hal-00954045, version 1

Citation

Achille Falaise, Agnès Tutin, Olivier Kraif, David Rouquet. ScienQuest: a treebank exploitation tool for non NLP-specialists. Conférence COLING 2012, 2012, Mumbai, India. 2012. 〈hal-00954045〉

Partager

Métriques

Consultations de la notice

83