AraStance: A Multi-Country and Multi-Domain Dataset of Arabic Stance Detection for Fact Checking

TL;DR

This paper introduces AraStance, a comprehensive Arabic stance detection dataset from multiple countries and domains, and benchmarks it with BERT models, highlighting its challenges and potential for improving fact-checking systems.

Contribution

The paper presents AraStance, a new large-scale, multi-domain Arabic stance detection dataset, and provides baseline results with BERT models to advance Arabic fact-checking research.

Findings

Best BERT model achieves 85% accuracy

Dataset covers diverse domains and countries

Stance detection remains a challenging task

Abstract

With the continuing spread of misinformation and disinformation online, it is of increasing importance to develop combating mechanisms at scale in the form of automated systems that support multiple languages. One task of interest is claim veracity prediction, which can be addressed using stance detection with respect to relevant documents retrieved online. To this end, we present our new Arabic Stance Detection dataset (AraStance) of 4,063 claim--article pairs from a diverse set of sources comprising three fact-checking websites and one news website. AraStance covers false and true claims from multiple domains (e.g., politics, sports, health) and several Arab countries, and it is well-balanced between related and unrelated documents with respect to the claims. We benchmark AraStance, along with two other stance detection datasets, using a number of BERT-based models. Our best model achieves an accuracy of 85\% and a macro F1 score of 78\%, which leaves room for improvement and reflects the challenging nature of AraStance and the task of stance detection in general.