PhD Dissertation Proposal by ZENG Fengzhu | Computational Fact-Checking With Limited Resources

The rapid dissemination of information through online platforms has sparked widespread concern about the propagation of misinformation. Manual fact-checking by professional fact-checkers is time-consuming and lacks scalability to address the vast volume of daily information. Consequently, computational fact-checking, driven by automated techniques in natural language processing (NLP), has garnered interest as a potential solution. However, computational fact-checking faces crucial challenges with limited resources: (1) the limited availability of features during the early stage of misinformation propagation; (2) the lack of training data for emerging events; (3) Large Language Models (LLMs) demand substantial computational resources.

To enhance computational fact-checking with limited resources, this thesis proposal focuses on developing methodologies that can make the most of the available data and computational resources, striking a balance between performance and resource availability. This thesis proposal introduces the stabilization process of early rumor detection to automatically determine when to make a timely, accurate and stable predictions with limited early-stage data. When limited labeled training instances are available, the thesis proposal proposes enabling a medium-size Pre-trained Language Model (PLM) to effectively compete with LLMs on evidence-based claim veracity prediction task. Finally, the thesis proposal presents the first solution for justification generation with evidence retrieval in the few-shot setting, using a medium-size PLM. Extensive experiments show that the proposed methodologies achieve more effective performance as compared to strong baselines, including LLMs. Lastly, this thesis proposal discusses some directions for future research in advancing computational fact-checking with limited resources.

ZENG Fengzhu is a Ph.D. candidate in Computer Science at SMU SCIS, supervised by Prof. GAO Wei. Her research interests intersect natural language processing and computational fact-checking.