Generalisability of Topic Models in Cross-corpora Abusive Language Detection - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2021

Generalisability of Topic Models in Cross-corpora Abusive Language Detection

Abstract

Rapidly changing social media content calls for robust and generalisable abuse detection models. However, the state-of-the-art supervised models display degraded performance when they are evaluated on abusive comments that differ from the training corpus. We investigate if the performance of supervised models for cross-corpora abuse detection can be improved by incorporating additional information from topic models, as the latter can infer the latent topic mixtures from unseen samples. In particular, we combine topical information with representations from a model tuned for classifying abusive comments. Our performance analysis reveals that topic models are able to capture abuse-related topics that can transfer across corpora, and result in improved generalisability.
Fichier principal
Vignette du fichier
14_Paper-5.pdf (328.29 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03212196 , version 1 (29-04-2021)

Identifiers

  • HAL Id : hal-03212196 , version 1

Cite

Tulika Bose, Irina Illina, Dominique Fohr. Generalisability of Topic Models in Cross-corpora Abusive Language Detection. NLP4IF 2021 - Workshop Censorship, Disinformation, and Propaganda, Jun 2021, Mexico city/Virtual, Mexico. ⟨hal-03212196⟩
186 View
253 Download

Share

Gmail Mastodon Facebook X LinkedIn More