Abstract:
In this study, we address the widespread issue of rumor propagation on social media. Current automated systems primarily rely on analyzing the stance made in tweets to predict the veracity of rumors. To enhance the effectiveness of these systems, we curated a new dataset using the Twitter API, annotating it for both claim and stance. Additionally, we extended our dataset by annotating well-known datasets such as Rumor-eval, incorporating claim annotations to emphasize the significance of stance alongside claims in detecting the veracity of the source tweet. Our approach till now involved a comprehensive literature review to understand existing methodologies and strategies in the field. We also implemented baseline models to evaluate their performance in achieving the same objective.
