Feature selection for drug-drug interaction detection using machine-learning based approaches

This paper describes the systems developed for the DDI Extraction challenge. The systems use machine learning methods and are based on SVM by using LIBSVM and SVMPerf tools. Classical features and corpus-specific features are selected according to their F-score. The best system obtained an F-measure of 0.5965.

Domaines

Informatique [cs] Informatique et langage [cs.CL]

Fichier principal

ddiextraction2011.pdf (253.98 Ko)

Origine	Fichiers produits par l'(les) auteur(s)

Brigitte Grau : Connectez-vous pour contacter le contributeur

https://hal.science/hal-02289969

Soumis le : mardi 17 septembre 2019-11:46:48

Dernière modification le : samedi 7 octobre 2023-21:36:20

Archivage à long terme le : samedi 8 février 2020-21:55:19

Dates et versions

hal-02289969 , version 1 (17-09-2019)

Identifiants

HAL Id : hal-02289969 , version 1

Citer

Anne-Lyse Minard, Lamia Makour, Anne-Laure Ligozat, Brigitte Grau. Feature selection for drug-drug interaction detection using machine-learning based approaches. Challenge Task on Drug-Drug Interaction Extraction (DDI), SEPLN, Sep 2011, Huelva, Spain. pp.43--50. ⟨hal-02289969⟩

Exporter

BibTeX XML-TEI Dublin Core DC Terms EndNote DataCite

Collections

CNRS LIMSI UNIV-PARIS-SACLAY SORBONNE-UNIVERSITE LISN ENSIIE

96 Consultations

67 Téléchargements