Stance Detection Dataset for Persian Tweets

Stance detection aims to identify an author's stance towards a specific topic which has become a critical component in applications such as fake news detection, claim validation, author profiling, etc. However, while the stance is easily detected by humans, machine learning models are falling short of this task. In the English language, due to having large and appropriate e datasets, relatively good accuracy has been achieved in this field, but in the Persian language, due to the lack of data, we have not made significant progress in stance detection. So, in this paper, we present a stance detection dataset that contains ۳۸۱۳ labeled tweets. We provide a detailed description of the newly created dataset and develop deep learning models on it. Our best model achieves a macro-average F۱-score of ۵۸%. Moreover, our dataset can facilitate research in some fields in Persian such as cross-lingual stance detection, author profiling, etc.