Skip to Main content Skip to Navigation
Conference papers

ExpLSA: An Approach Based on Syntactic Knowledge in Order to Improve LSA for a Conceptual Classification Task

Nicolas Béchet 1 Jacques Chauché 2, 1 Mathieu Roche 3, 1
1 TEXTE - Exploration et exploitation de données textuelles
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : Latent Semantic Analysis (LSA) is nowadays used in various thematic like cognitive models, educational applications but also in classification. We propose in this paper to study different methods of proximity of terms based on LSA. We improve this semantic analysis with additional semantic information using Tree-tagger or a syntactic analysis to expand the studied corpus. We finally apply LSA on the new expanded corpus.
Complete list of metadata

https://hal-lirmm.ccsd.cnrs.fr/lirmm-00335879
Contributor : Nicolas Béchet <>
Submitted on : Thursday, October 30, 2008 - 7:33:39 PM
Last modification on : Tuesday, June 15, 2021 - 4:19:34 PM
Long-term archiving on: : Tuesday, October 9, 2012 - 2:43:28 PM

File

Cicling08_NB_et_al.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : lirmm-00335879, version 1

Citation

Nicolas Béchet, Jacques Chauché, Mathieu Roche. ExpLSA: An Approach Based on Syntactic Knowledge in Order to Improve LSA for a Conceptual Classification Task. CICLing: Conference on Intelligent Text Processing and Computational Linguistics, Feb 2008, Haifa, Israel. pp.213-224. ⟨lirmm-00335879⟩

Share

Metrics

Record views

317

Files downloads

275