IRIT at TRAC 2018 - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2018

IRIT at TRAC 2018

Résumé

This paper describes the participation of the IRIT team to the TRAC 2018 shared task on Aggres-sion Identification and more precisely to the shared task in English language. The three followingmethods have been used: a) a combination of machine learning techniques that relies on a setof features and document/text vectorization, b) Convolutional Neural Network (CNN) and c) acombination of Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM).Best results were obtained when using the method (a) on the English test data from Facebookwhich ranked our method sixteenth out of thirty teams, and the method (c) on the English testdata from other social media, where we obtained the fifteenth rank out of thirty.

Mots clés

Fichier principal
Vignette du fichier
ramiandrisoa_26318.pdf (138.07 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02942340 , version 1 (17-09-2020)

Identifiants

  • HAL Id : hal-02942340 , version 1
  • OATAO : 26318

Citer

Faneva Ramiandrisoa, Josiane Mothe. IRIT at TRAC 2018. 1st Workshop on Trolling, Aggression and Cyberbullying, in International @ Conference of Computational Linguistics (TRAC@COLING 2018), Aug 2018, Santa Fe, New Mexico, United States. pp.19-27. ⟨hal-02942340⟩
20 Consultations
48 Téléchargements

Partager

Gmail Facebook X LinkedIn More