The role of claims in misinformation detection
(2024) 120th APSA Annual Meeting & Exhibition — Location: Philadelphia (5.September.2024)
Files
No attached file found for this publication.
Details
- Authors
- Abstract
- Misinformation in the digital age poses a significantly greater threat to democratic processes than its analog predecessors, given its decentralized nature and ability to propagate with unprecedented speed and precision. This evolution challenges traditional methods of detection and mitigation, amplifying its potential to erode public trust, distort electoral outcomes, and intensify political polarization. The World Economic Forum's 2024 global risk survey underscores the urgency, ranking online misinformation as the most concerning imminent global risk. While human-centric interventions such as boosting, nudging, and debunking aim to enhance users' critical evaluation of information, technological approaches like automated content labeling offer scalable and rapid solutions, albeit with limitations tied to method effectiveness. This paper contributes to advancing technology-centric approaches by focusing on the automated identification of claims—assertions of truth—within vast social media content. This critical element forms part of a broader framework that leverages network properties, linguistic features, propagation patterns, and aggregate user behaviors to distinguish true claims from false ones. The paper begins by evaluating existing misinformation detection strategies, highlighting their strengths and weaknesses. It then builds upon unsupervised frameworks, emphasizing the value of focusing on claims to improve detection. A method for identifying claims within social media posts is proposed and demonstrated through an application involving COVID-related posts on X (formerly Twitter) from the early pandemic period. The paper concludes by discussing future steps for extending this approach within the larger framework for misinformation detection.
- Affiliations
Citations
Clifton van der Linden, Deena Abul-Fottouh, Hugo Mailhot, Timmers, C., Vande Kerckhove, C., & John R. McAndrews. (2024). The role of claims in misinformation detection. 120th APSA Annual Meeting & Exhibition, Philadelphia. https://hdl.handle.net/2078.5/260906