Sentiment Analysis on YouTube Comments Using YouTube Data API V3 | IJCT Volume 13 – Issue 2 | IJCT-V13I2P51
International Journal of Computer Techniques
ISSN 2394-2231
Author
Bhuvansh Hiwrekar, Saksham Gorade, Shripad Bhakre, Tejas Wath, Shritesh Dharmale, Prof. Pravin Kaware
Abstract
In the era of digital media, YouTube has emerged as one of the most influential platforms for sharing information, entertainment, and public opinion. With billions of daily interactions, the platform generates enormous volumes of user-generated textual data in the form of comments. Understanding audience sentiment from these comments holds significant value for content creators, marketers, brand managers, and researchers. This paper presents a comprehensive study on automated Sentiment Analysis of YouTube comments using the YouTube Data API v3 for data collection and Natural Language Processing (NLP) techniques for classification. The proposed system fetches comment data from target YouTube videos, preprocesses the raw text, and employs multiple machine learning and deep learning models — including VADER, TextBlob, Support Vector Machine (SVM), Naive Bayes, and transformer-based BERT — to classify sentiments into Positive, Negative, and Neutral categories. Experimental results demonstrate that BERT achieves the highest classification accuracy of 93.1%, outperforming all traditional baselines. The system further provides interactive visual dashboards and trend reports to support data-driven decision-making.
Keywords
BERT, Deep Learning, Machine Learning, Natural Language Processing, Opinion Mining, Sentiment Analysis, Social Media Mining, YouTube Data API v3.
Conclusion
This paper presented a comprehensive, end-to-end automated Sentiment Analysis system for YouTube comments, integrating the YouTube Data API v3 with a multi-stage NLP preprocessing pipeline and a rigorous comparative evaluation of five classification models. Experimental results on 45,000 YouTube comments across five content categories demonstrate that the fine-tuned BERT transformer model achieves the highest overall performance (93.1% accuracy, 92.9% F1-score), significantly outperforming lexicon-based baselines and traditional ML classifiers. The system’s interactive visual dashboards make sentiment insights accessible to non-technical users, enabling practical deployment across content creation, brand monitoring, and research applications. [6], [7], [4], [1], [3], [8], [13]
Future work directions include: (1) incorporating dedicated sarcasm and irony detection modules via multi-task learning, (2) extending to multilingual analysis using mBERT, (3) integrating aspect-based sentiment analysis for finer-grained feedback extraction, (4) real-time streaming comment analysis using Apache Kafka for live video events, and (5) incorporating multimodal signals such as reply thread context and video metadata to further improve classification accuracy. [5], [14], [16], [18]
References
[1] S. Qin, R. Menezes, and M. Silaghi, “A Recommender System for YouTube Based on its Network of Reviewers,” in Proc. IEEE Second International Conference on Social Computing, 2010. [2] T. O’Keefe and I. Koprinska, “Feature Selection and Weighting in Sentiment Analysis,” in Proc. 14th Australasian Document Computing Symposium, Sydney, Australia, December 2009. [3] W. Yang and Z. Qian, “Understanding the Characteristics of Category-Specific YouTube Videos,” December 2011, Available: CiteSeerX. [4] A. N. Muhammad, S. Bukhori, and P. Pandunata, “Sentiment Analysis of Positive and Negative of YouTube Comments Using Naïve Bayes–Support Vector Machine (NBSVM) Classifier,” in Proc. International Conference on Computer Science, Information Technology, and Electrical Engineering (ICOMITEE), 2019. [5] K. P. Gunasekaran, “Exploring Sentiment Analysis Techniques in Natural Language Processing: A Comprehensive Review,” arXiv preprint arXiv:2305.14842, May 2023. [6] S. Al Osaimi and K. M. Badruddin, “Sentiment Analysis of Arabic Tweets Using RapidMiner,” Dept. of Information System, Imam Muhammad ibn Saud Islamic University, KSA. [7] P. Tripathi, S. K. Vishwakarma, and A. Lala, “Sentiment Analysis of English Tweet Using Rapidminer,” in Proc. International Conference on Computational Intelligence and Communication Networks, 2015, pp. 668–672. [8] D. A. Musleh, I. Alkhwaja, A. Alkhwaja, M. Alghamdi, H. Abahussain, F. Alfawaz, N. Min-Allah, and M. M. Abdulqader, “Arabic Sentiment Analysis of YouTube Comments: NLP-Based Machine Learning Approaches for Content Evaluation,” Big Data and Cognitive Computing, MDPI, vol. 7, no. 3, p. 127, July 2023. DOI: 10.3390/bdcc7030127. [9] S. Pichad, S. Kamble, R. Kalamb, and S. Chavan, “Analysing Sentiments for YouTube Comments Using Machine Learning,” International Journal for Research in Applied Science & Engineering Technology (IJRASET), vol. 11, issue V, May 2023. [10] R. Giri, M. Sirsath, and H. T. Kanakia, “Youtube Comments Sentiment Analysis,” in Proc. IEEE 9th International Conference for Convergence in Technology (I2CT), 2024, pp. 1–4. [11] A. S. Aiswarya and H. Rajeev, “Youtube Comment Sentimental Analysis,” Indian Journal of Data Mining (IJDM), vol. 4, issue 1, May 2024. DOI: 10.2139/ssrn.4846213. [12] J. L. Rachel, M. Sawood, A. Khan, and B. P. Kumar, “Extracting Sentiments from YouTube Comments Using Deep Learning and Transformers,” International Journal for Research in Applied Science & Engineering Technology (IJRASET), 2024. DOI: 10.22214/ijraset.2023.57783. [13] J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” in Proc. NAACL-HLT 2019, pp. 4171–4186. DOI: 10.18653/v1/N19-1423. [14] N. Nahas, P. Swetha, and R. Nandakumar, “Classifiers for Sentiment Analysis of YouTube Comments: A Comparative Study,” in Proc. International Conference on Machine Learning, Deep Learning and Computational Intelligence for Wireless Communication (MDCWC 2023), Springer, 2024. DOI: 10.1007/978-3-031-47942-7_38. [15] “Transformer-based models for sentiment analysis of YouTube video comments,” ScienceDirect, 2025. DOI: 10.1016/j.array.2025.x. [16] A. Shevtsov et al., “TWSSenti: A Novel Hybrid Framework for Topic-Wise Sentiment Analysis on Social Media Using Transformer Models,” arXiv preprint arXiv:2504.09896, April 2025. [17] H. Malik and Z. Tian, “A Framework for Collecting YouTube Meta-Data,” Procedia Computer Science, vol. 113, pp. 194–201, 2017. DOI: 10.1016/j.procs.2017.08.347. [18] P. Surana, S. Vispute, M. Kalbhor, P. Ugale, I. Syeda, M. Yakumsha, and A. Suryawanshi, “Opinion Mining of YouTube Video Comments Using Machine Learning and Deep Learning,” Springer, 2026. DOI: 10.1007/978-3-032-06694-7_13. [19] K. Wankhade, A. C. S. Rao, and C. Kulkarni, “A Survey on Sentiment Analysis Methods, Applications, and Challenges,” Artificial Intelligence Review, Springer, vol. 55, no. 7, pp. 5731–5780, 2022. [20] B. Pang, L. Lee, and S. Vaithyanathan, “Thumbs Up? Sentiment Classification Using Machine Learning Techniques,” in Proc. EMNLP 2002, pp. 79–86.
How to Cite This Paper
Bhuvansh Hiwrekar, Saksham Gorade, Shripad Bhakre, Tejas Wath, Shritesh Dharmale, Prof. Pravin Kaware (2026). Sentiment Analysis on YouTube Comments Using YouTube Data API V3. International Journal of Computer Techniques, 13(2). ISSN: 2394-2231.
