Sentiment Analysis for YouTube Comments and Videos Using AI
1Dr.P. Rajendra Prasad
Associate Professor, Department of Computer Science and Engineering Vignan’s Institute of Management and Technology for Women, Hyd.
Email:rajipe@gmail.com
2Valluri Poojaprasanna
UG Student, Department of Computer Science and Engineering Vignan’s Institute of Management and Technology for Women, Hyd.
Email:poojaprasannav@gmail.com
3Atmakur Bhuvaneshwari
UG Student, Department of Computer Science and Engineering5 Vignan’s Institute of Management and Technology for Women, Hyd.
Email: bhuvaneshwariatmakur@gmail.com
4Bandla Swapna
UG Student, Department of Computer Science and Engineering Vignan’s Institute of Management and Technology for Women Email:swapnabandla14@gmail.com
5G. Sai Yamini Devi
UG Student, Department of Computer Science and Engineering Vignan’s Institute of Management and Technology for Women, Hyd.
Email: yamireddy2004@gmail.com
Abstract- The YouTube Comments and Videos Sentiment Analysis project is an advanced system that automates the sentiment classification of both comments and video transcripts associated with YouTube videos. Using Natural Language Processing (NLP) and cutting-edge deep learning models, the system determines whether the sentiments expressed are positive, negative, or neutral. The system integrates tools such as the YouTube Data API to extract comments and the YouTube Transcript API to retrieve video transcripts. Advanced pre-processing techniques like tokenization, lemmatization, and stopword removal are applied to prepare the text for sentiment analysis. The project employs pre-trained models, including BERT for comments sentiment analysis and LSTM and GRU for video transcript sentiment analysis, ensuring high accuracy in sentiment classification. Users interact with the system through a web-based interface where they can input a YouTube video URL. The system then retrieves comments and transcripts, performs sentiment analysis, and visualizes the results in an intuitive graphical format. To ensure the system delivers reliable results, various performance indicators are used to evaluate how well it identifies and interprets sentiment. These measures help verify the system’s effectiveness and consistency. This project aims to offer real-time analysis of public opinion, providing valuable insights to content creators, marketers, and businesses. As a result, they can make smarter and faster decisions based on current trends and audience reactions.
Keywords- Sentiment, YouTube Comments, YouTube URL, LSTM, GRU, BERT, NLP.
