A Framework for Identifying Clickbait

A Framework for Identifying Clickbait

Adithya A J*, Krishna J#, Rona Siby&, Roshan Babu@, Mary Priyanka K S+
Department of CSE, College of Engineering Kidangoor, Kottayam, Kerala, India

* aadianeesh@gmail.com, #kj163705@gmail.com, &ronasiby03@gmail.com,
@roshanbabu086@gmail.com, +marypriyanka@ce-kgr.org

Abstract—The pervasive use of clickbait in YouTube videos negatively impacts user experience by leveraging attention-grabbing headlines, descriptions, and thumbnails to attract views. This deceptive practice has prompted significant research into automated detection frameworks aimed at mitigating its effects. Existing methods for clickbait detection primarily utilize machine learning and deep learning techniques, with approaches focusing on text-based metadata analysis, visual pattern recognition in thumbnails, and sentiment analysis of video titles and descrip­tions. Transformer models have shown promise in analyzing complex patterns in both textual and visual data, while sentiment analysis provides contextual validation for identifying misleading content. While considerable progress has been made, further research is needed to enhance the robustness and scalability of these frameworks in addressing the dynamic nature of clickbait strategies.

Index Terms—Deep learning, Machine learning, clickbait de­tection