Sentimental Analysis of YouTube Video Comments Using Bagging Ensemble Learning Approach

Abstract

An important indicator that shows how well-liked a YouTube video is by its viewers is the like ratio. By examining the emotive tone of viewer comments, sentiment analysis can be used to forecast the like ratio of a YouTube video. With this method, the YouTube API is used to first get the comments from the video. Following that, these comments are pre-processed to eliminate any unnecessary data, including URLs and special characters, and to change the text's case to lowercase. The pre-processed comments are then subjected to sentiment analysis using a natural language processing package, such as TextBlob or NLTK, to categorise them as positive, negative, or neutral. The like ratio can be estimated after sentiment analysis by measuring the percentage of positive comments to all comments. This can be used to determine how viewers feel about the video overall and forecast whether the film will have a high or low like ratio. Overall, forecasting the like ratio of a YouTube video using sentiment analysis can offer insightful information for content producers and marketers, assisting them in understanding the emotional response of their audience and improving their content accordingly. KEYWORDS: Text mining, Sentimental Analysis, Youtube, NLTK, Machine Learing Processing