cantnlp@LT-EDI-2024: Automatic Detection of Anti-LGBTQ+ Hate Speech in Under-resourced Languages

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wong, Sidney G. -J., Durward, Matthew
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910310107971584
author Wong, Sidney G. -J.
Durward, Matthew
author_facet Wong, Sidney G. -J.
Durward, Matthew
contents This paper describes our homophobia/transphobia in social media comments detection system developed as part of the shared task at LT-EDI-2024. We took a transformer-based approach to develop our multiclass classification model for ten language conditions (English, Spanish, Gujarati, Hindi, Kannada, Malayalam, Marathi, Tamil, Tulu, and Telugu). We introduced synthetic and organic instances of script-switched language data during domain adaptation to mirror the linguistic realities of social media language as seen in the labelled training data. Our system ranked second for Gujarati and Telugu with varying levels of performance for other language conditions. The results suggest incorporating elements of paralinguistic behaviour such as script-switching may improve the performance of language detection systems especially in the cases of under-resourced languages conditions.
format Preprint
id arxiv_https___arxiv_org_abs_2401_15777
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle cantnlp@LT-EDI-2024: Automatic Detection of Anti-LGBTQ+ Hate Speech in Under-resourced Languages
Wong, Sidney G. -J.
Durward, Matthew
Computation and Language
This paper describes our homophobia/transphobia in social media comments detection system developed as part of the shared task at LT-EDI-2024. We took a transformer-based approach to develop our multiclass classification model for ten language conditions (English, Spanish, Gujarati, Hindi, Kannada, Malayalam, Marathi, Tamil, Tulu, and Telugu). We introduced synthetic and organic instances of script-switched language data during domain adaptation to mirror the linguistic realities of social media language as seen in the labelled training data. Our system ranked second for Gujarati and Telugu with varying levels of performance for other language conditions. The results suggest incorporating elements of paralinguistic behaviour such as script-switching may improve the performance of language detection systems especially in the cases of under-resourced languages conditions.
title cantnlp@LT-EDI-2024: Automatic Detection of Anti-LGBTQ+ Hate Speech in Under-resourced Languages
topic Computation and Language
url https://arxiv.org/abs/2401.15777