Skip to Main content Skip to Navigation
Conference papers

Apport des dépendances syntaxiques et des patrons séquentiels à l'extraction de relations

Abstract : In this paper, we investigate the contribution of syntactic features to the task of unsupervised clustering of semantic relation instances. Instances, i.e. couples of concepts appearing in scientific texts, are represented in a couple-pattern matrix over co-occurrence contexts. Various possible contextual representation features are compared, using sequential pattern mining and syntactic path extraction. We compare the purely lexical feature space with a combined representation, and conclude that adding syntactic features has the potential to improve clustering performance.
Complete list of metadatas

Cited literature [40 references]  Display  Hide  Download

https://hal.inria.fr/hal-02079719
Contributor : Kata Gabor <>
Submitted on : Tuesday, March 26, 2019 - 11:25:31 AM
Last modification on : Tuesday, January 5, 2021 - 5:28:07 PM
Long-term archiving on: : Thursday, June 27, 2019 - 2:31:23 PM

File

TALN2018__Apport_des_d_pendanc...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02079719, version 1

Citation

Kata Gábor, Nadège Lechevrel, Isabelle Tellier, Thierry Charnois, Haifa Zargayouna, et al.. Apport des dépendances syntaxiques et des patrons séquentiels à l'extraction de relations. TALN 2018, May 2018, Rennes, France. ⟨hal-02079719⟩

Share

Metrics

Record views

200

Files downloads

181