Identification of Clickbaits from Video Sharing Platforms
People nowadays spend most of their time learning or entertaining themselves on video-sharing platforms. These platforms (e.g., YouTube, Vimeo) have already begun to supplant television day by day. However, some content creators use irrelevant and eye-catching titles (clickbait) that do not correspond to the video's original content. This video title increases video engagement or view counts and ad revenue but wastes viewers' valuable time. These videos are spreading like a contagion in the video-sharing platforms. In this research, we have targeted to solve this issue by identifying clickbait videos by creating a dataset containing video titles (Bangla, English, and Phonetic Bangla) and numerical features. We have adopted various feature extraction techniques for proper analysis and applied Logistic Regression, Support Vector Machine, Decision Tree, and Gradient Boost algorithm to create classification models. Remarkably, both the Decision Tree and Gradient Boost models have accomplished a 99% accuracy score in detecting clickbait videos.