Topic Modeling and Sentiment Analysis of YouTube Comments on OpenAI’s Stargate Project Using BERTopic and RoBERTa
DOI:
https://doi.org/10.32664/j-intech.v14i03.2395Kata Kunci:
BERTopic, Data Center, OpenAI, RoBERTa, Sentiment Analysis, Stargate, Topic ModelingAbstrak
The rapid development of generative artificial intelligence has sparked growing interest, widespread adoption across various sectors, and intense public debate, particularly following the announcement of the Stargate project by OpenAI, SoftBank, and Oracle. The diverse public reactions documented on social media platforms, particularly in YouTube comment sections, can serve as a valuable source of public opinion data. The YouTube comment data used in this study was collected through scraping and subsequently processed through a cleaning procedure, yielding clean data covering the period from January 21, 2025, to April 21, 2026. This study aims to identify dominant discussion topics and the underlying sentiment to uncover indications of public concerns and expectations regarding the AI development project led by OpenAI through the Stargate initiative. The analysis in this study employs a structured NLP approach, specifically BERTopic for topic modeling and RoBERTa for sentiment analysis. The RoBERTa model was fine tuned for binary classification using a combination of pseudo labeling, annotation, class balancing, and classification decision threshold optimization, which optimally yielded an accuracy of 0.89 and a macro F1 score of 0.87 in the two class evaluation. BERTopic successfully identified 43 topics with a coherence score of 0.58 and a topic diversity of 0.71.
Referensi
[1] I. Zaenudin and A. B. Riyan, “Perkembangan Kecerdasan Buatan (AI) Dan Dampaknya Pada Dunia Teknologi,” Jitu: Jurnal Informatika Utama, vol. 2, no. 2, pp. 128–153, 2024, doi: https://doi.org/10.55903/jitu.v2i2.240.
[2] V. C. Storey, W. T. Yue, J. L. Zhao, and R. Lukyanenko, “Generative Artificial Intelligence: Evolving Technology, Growing Societal Impact, and Opportunities for Information Systems Research,” Information Systems Frontiers, vol. 27, no. 5, pp. 2081–2102, Oct. 2025, doi: 10.1007/s10796-025-10581-7.
[3] ClearML, AI Infrastructure Alliance, and Furiosa, “The State of AI Infrastructure at Scale 2024,” 2024.
[4] OpenAI, “Announcing The Stargate Project.” Accessed: Feb. 20, 2026. [Online]. Available: https://openai.com/id-ID/index/announcing-the-stargate-project/
[5] Bloomberg Originals, “Inside OpenAI’s Stargate Megafactory with Sam Altman.” Accessed: Jan. 30, 2026. [Online]. Available: https://youtu.be/GhIJs4zbH0o?si=xN2be82cHkiAGMpn
[6] A. Ocal, “Perceptions of the Future of Artificial Intelligence on Social Media : A Topic Modeling and Sentiment Analysis Approach,” IEEE Access, vol. 12, no. December, 2024, doi: 10.1109/ACCESS.2024.3510526.
[7] W. Wahyuni, T. P. Lestari, M. Apriliana, and R. Gumelta, “Implementation of BERTopic for Topic Modeling Analysis of the Free Nutritious Meal Program Based on YouTube Comments,” 2025.
[8] B. Rahmadhani, Purwono, and S. D. Kurniawan, “Understanding Transformers: A Comprehensive Review,” Journal of Advanced Health Informatics Research (JAHIR), vol. 2, no. 2, pp. 85–94, 2024, doi: 10.59247/jahir.v2i2.292.
[9] A. Söderwall and G. Telešova, “A Comparative Study with LDA and BERTopic : AI Policies Across Different Democracy Indexes,” 2025.
[10] M. David Angelo, R. Widyatna Harwenda, I. Budi, A. Budi Santoso, and P. Kresna Putra, “Sentiment Analysis and Topic Modeling of Public Opinion on Indonesia New Capital City Development Policies,” Journal Eduvest, vol. 5, no. 5, pp. 5951–5969, 2025.
[11] A. Alamsyah and N. D. Girawan, “Improving Clothing Product Quality and Reducing Waste Based on Consumer Review Using RoBERTa and BERTopic Language Model,” Big Data and Cognitive Computing, vol. 7, 2023, doi: 10.3390/bdcc7040168.
[12] A. Vishnevskaya, V. Pavlopoulos, and J. Stott, “Which Model Feels Better ? A Comparison of Computational Approaches to Emotion Detection in Social Media with Imbalanced Data,” Informatics, pp. 1–24, 2026.
[13] A. S. Aribowo, S. Khomsah, and S. Saifullah, “Semi-Supervised Sentiment Classification Using Self-Learning and Enhanced Co-Training,” JURNAL INFOTEL, pp. 472–489, 2025, doi: https://doi.org/10.20895/infotel.v17i3.1344.
[14] A. K. Jayaraman, A. Murugappan, T. E. Trueman, G. Ananthakrishnan, and A. Ghosh, “Imbalanced aspect categorization using bidirectional encoder representation from transformers,” Procedia Comput. Sci., vol. 218, pp. 757–765, 2022, doi: 10.1016/j.procs.2023.01.056.
[15] L. Bojić et al., “Comparing large Language models and human annotators in latent content analysis of sentiment , political leaning , emotional intensity and sarcasm.” Sci. Rep., pp. 1–16, 2025, doi: 10.1038/s41598-025-96508-3.
[16] H. N. Alfiana, A. Doewes, and B. Widoyono, “Aspect-Based Sentiment Analysis of Access by KAI Application Reviews Using IndoBERT for Multi-Label Classification Tasks,” Jurnal Teknik Informatika (JUTIF), vol. 7, no. 1, pp. 286–306, 2026, doi: 10.52436/1.jutif.2026.7.2.5402 Aspect-Based.
[17] W. J. Kusoema, I. Ibrahim, P. Studi, T. Informatika, K. Bandung, and J. Barat, “Analisis Sentimen dalam Kasus Korupsi PT . Pertamina menggunakan Metode indoBERT dan RCNN,” Sistemasi: Jurnal Sistem Informasi, vol. 14, pp. 2246–2257, 2025.
[18] S. Si, Y. Gao, H. Sun, Y. Zhang, and H. Luo, “Revisiting the Role of Label Smoothing in Enhanced Text Sentiment Classification,” Electronics, pp. 1–15, 2026, doi: https://doi.org/10.3390/electronics15101984.
[19] M. Abdelhamid and A. Desai, “Balancing the Scales : A Comprehensive Study on Tackling Class Imbalance in Binary Classification,” 2024.
[20] J. Hancock, J. M. Johnson, and T. M. Khosgoftaar, “A Comparative Approach to Threshold Optimization for Classifying Imbalanced Data,” in 2022 IEEE 8th International Conference on Collaboration and Internet Computing (CIC), IEEE (Institute of Electrical and Electronics Engineers), 2022, pp. 135–142. doi: 10.1109/CIC56439.2022.00028.
[21] N. Varshney, S. Mishra, and C. Baral, “Towards Improving Selective Prediction Ability of NLP Systems,” in 7th Workshop on Representation Learning for NLP, RepL4NLP 2022 at ACL 2022, Dublin: Association for Computational Linguistics (ACL), 2022, pp. 221–226.
[22] B. D. Jesudas et al., “Utilizing BERTopic Modeling for Concept Discovery in the Domain of Gerotranscendence and Solitude,” Res. Sq., pp. 1–23, 2025, doi: 10.21203/rs.3.rs-7383440/v1.
[23] C. Schneijderberg, O. Wieczorek, and I. Steinhardt, THE HANDBOOK OF QUALITATIVE AND QUANTITATIVE CONTENT ANALYSIS. Taylor & Francis Group, 2026. doi: https://doi.org/10.4324/9781003496397.
[24] M. Grootendorst, “BERTopic: Neural topic modeling with a class-based TF-IDF procedure,” Mar. 2022.
[25] A. F. A. Mahmoud, F. A. Abdalla, G. Saad, M. Khamis, and M. A. Hamed, “A Comparative Evaluation of LDA, NMF, and BERTopic: Analyzing Perplexity and Coherence Metrics,” International Information and Engineering Technology Association, vol. 30, no. 12, pp. 3163–3169, 2025.
[26] A. Alharbi, A. A. Hai, R. Aljurbua, and Z. Obradovic, “AI-Driven Sentiment Trend Analysis: Enhancing Topic Modeling Interpretation with ChatGPT,” IFIP Adv. Inf. Commun. Technol., vol. 712, no. Ml, pp. 3–17, 2024, doi: 10.1007/978-3-031-63215-0_1.
[27] S. Samsir, R. S. Saragih, S. Subagio, R. Aditiya, and R. Watrianthos, “BERTopic Modeling of Natural Language Processing Abstracts: Thematic Structure and Trajectory,” Jurnal Media Informatika Budidarma, vol. 7, no. 3, p. 1514, 2023, doi: 10.30865/mib.v7i3.6426.
Unduhan
Diterbitkan
Terbitan
Bagian
Lisensi
Hak Cipta (c) 2026 J-INTECH ( Journal of Information and Technology)
Artikel ini berlisensiCreative Commons Attribution-ShareAlike 4.0 International License.
