MGKM at StanceEval2024 Fine-Tuning Large Language Models for Arabic Stance Detection
Mamoun Alghaslan, Khaled Almutairy · 2024
Social media platforms have become essential in daily life, enabling users to express their opinions and stances on various topics.Stance detection is a task that identifies the viewpoint expressed in text toward a target subject.Despite the growing importance of Arabic tweets in shaping public opinion, there is a lack of research on stance detection in this domain.In this work, we evaluate the effectiveness of fine-tuning three Large Language Models (LLMs) in detecting target-specific stances in the MAWQIF dataset (Alturayeif et al., 2022).The LLMs assessed are ChatGPT-3.5turbo,Meta-Llama-3-8B-Instruct, and Falcon-7B-Instruct.Our findings demonstrate that finetuning substantially enhances the stance detection capabilities of LLMs in Arabic tweets.GPT-3.5-Turboexhibits the highest performance among the evaluated models, achieving a macro-F1 score of 82.93.Our work ranked second in the StanceEval2024 leaderboard, on a blind test set (Alturayeif et al., 2024).