Evaluating YouTube Video via Sentiment Analysis: A Case Study in Code-Mixed Bangla-English Context

This study examines YouTube videos that incorporate code-mixed Bangla-English content in their comment section through sentiment analysis. With the proliferation of code-mixed content on social media platforms, understanding the sentiment expressed towards such videos becomes crucial. This research investigates the feasibility and effectiveness of sentiment analysis techniques tailored for code-mixed text in social media. YouTube is the largest online repository of videos and it provides a comment section to enable viewers to express their opinions. This comment section can be a good source for analysis which can potentially lead to the development of many critical applications. This paper introduces a novel technique to rate YouTube videos utilizing a Bangla sentiment analysis model built on the state-of-the-art transformer architecture. Our approach links viewers’ sentiments to a useful rating system, effectively empowering the users to understand YouTube videos.