Fake news detection using passive aggressive classifier
M. Nagaraju, S. Sudheer Reddy, M. Arunkumar, Pothumarthi Sridevi, Pula Sekhar, Natha Deepthi · 2025
Social Media is full of applications for fake news, including trend analysis, advertisement, and recommendation systems. Its use in predicting the popularity of a particular topic is challenging due to the factors that influence its quality and relevance to viewers. To treat them equally, other methods usually include numerous factors and modalities in their models. In this paper, we present a technique that combines the numerical and semantic features to improve the performance of a post&s;s popularity prediction. The proposed method utilizes a self-attention mechanism to solve this issue. The paper comprehensively evaluates the proposed method, which considers the various aspects of the ACM Media SMPD-2020 dataset. The results of the evaluation compare the proposed approach with other methods.