Stance Classification in Social Media Using Machine Learning Techniques

Stance classification has been a popular research topic. The target of my research is to build a model to classify post stance in social media, which can later be used to determine the veracity of rumors in social media. My research mainly consists four parts: related work review, baseline reproduction, model exploration and optimization, and results and analysis. All the posts are classified into four categories: support, deny, query and comment. I first tried to reproduce the baseline. Then, I explored and evaluated the performances of different models, and tried to improve the model by adding new features. Finally, I summarized the results and gave future research directions.