Hybrid NLP framework for enhanced sentiment analysis and topic detection on YouTube
Ramesh Babu, Sri Ramakrishna, Suneel Kumar Duvvuri · International Journal of Basic and Applied Sciences · 2025
This paper introduces an advanced hybrid NLP framework designed to enhance sentiment analysis and topic detection in YouTube comments. By combining feature extraction methods like Bag of Words (BoW) and TF-IDF with neural network models like LSTM and Bi-LSTM, the framework effectively uncovers latent topics and sentiment orientations. The study specifically analyzes comments on Oscar-nominated movie trailers, demonstrating the framework's ability to capture both explicit and implicit patterns of sentiment. This study shows that the Bi-LSTM model with BoW features achieves the highest performance, with accuracy, precision, recall, and F1-scores nearing 90%. This hybrid approach delivers practical and theoretical developments in natural language processing applications for content creators and marketers to optimize engagement strategies based on user sentiment and thematic preferences.