Hybrid NLP framework for enhanced sentiment analysis and topic detection on YouTube
International Journal of Basic and Applied Sciences · 14.05.2025 · 14 (1) · 304-313
DOI: 10.14419/kft8ae18 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Basic and Applied Sciences, online erschienen am 14.05.2025. Verfasst von Ramesh Babu und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref).
Abstract (Originalsprache)
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.
Zitieren
APA
Babu, R., Ramakrishna, S., & Duvvuri, S. K. (2025). Hybrid NLP framework for enhanced sentiment analysis and topic detection on YouTube. International Journal of Basic and Applied Sciences, 14(1), 304-313. https://doi.org/10.14419/kft8ae18
Vancouver
Babu R, Ramakrishna S, Duvvuri SK. Hybrid NLP framework for enhanced sentiment analysis and topic detection on YouTube. International Journal of Basic and Applied Sciences. 2025;14(1):304-313. doi:10.14419/kft8ae18
BibTeX
@article{Babu2025,
author = {Babu, Ramesh and Ramakrishna, S. and Duvvuri, Suneel Kumar},
title = {Hybrid NLP framework for enhanced sentiment analysis and topic detection on YouTube},
journal = {International Journal of Basic and Applied Sciences},
year = {2025},
volume = {14},
number = {1},
pages = {304--313},
doi = {10.14419/kft8ae18}
}
EUVIMED