Communication Dans Un Congrès Année : 2017

Unsupervised Event Clustering and Aggregation from Newswire and Web Articles

Résumé

In this paper, we present an unsupervisedpipeline approach for clustering news articlesbased on identified event instances intheir content. We leverage press agencynewswire and monolingual word alignmenttechniques to build meaningful andlinguistically varied clusters of articlesfrom the Web in the perspective of abroader event type detection task. We validateour approach on a manually annotatedcorpus of Web articles.

Fichier principal
Vignette du fichier
W17-4211 (234.98 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

cea-01857885 , version 1 (17-08-2018)

Licence

Identifiants

Citer

Swen Ribeiro, Olivier Ferret, Xavier Tannier. Unsupervised Event Clustering and Aggregation from Newswire and Web Articles. 2017 EMNLP Workshop: Natural Language Processing meets Journalism, 2017, Copenhagen, Denmark. pp.62-67, ⟨10.18653/v1/W17-4211⟩. ⟨cea-01857885⟩
295 Consultations
281 Téléchargements

Altmetric

Partager

  • More