Humans Need Context, What About Machines? Investigating Conversational Context in Abusive Language Detection
COM3 Level 1
SR14, COM3 01-23
Abstract:
A crucial aspect of abusive language on social media platforms (toxicity, hate speech, harmful stereotypes, etc.) is its inherent contextual nature. In this talk, we focus on the role of conversational context in abusive language detection, one of the most “direct” forms of context in this domain, as given by the conversation threads (e.g., directly preceding message, original post). The incorporation of surrounding messages has proven vital for the accurate human annotation of harmful content. However, many prior works have either ignored this aspect, collecting and processing messages in isolation or have obtained inconsistent results when attempting to embed such contextual information into traditional classification methods. The reasons behind these findings have not yet been properly addressed. To this end, we propose an analysis of the impact of conversational context in abusive language detection together with two alternative context-aware architectures.
Reference:
Tom Bourgeade, Zongmin Li, Farah Benamara, Véronique Moriceau, Jian Su and Aixin Sun. Humans Need Context, What About Machines? Investigating Conversational Context in Abusive Language Detection. To appear in the proceedings of LREC-COLING 2024
Bio
Farah Benamara is a Full Professor in Computer Science at Paul Sabatier University in Toulouse, France. She is a member of the Artificial Intelligence Department of the IRIT laboratory. She is a co-head of MELODI (MEthodes et ingénierie des Langues, des Ontologies et du DIscours), a research group in NLP, Ontologies and Semantic Web technologies. Since September 2023, Farah joined CNRS@CREATE in Singapore. She is also a member of IPAL, a CNRS International (French-Singaporean) Research Laboratory on Artificial Intelligence (IPAL) where she co-lead the NLP theme.