Implementasi BERT dan IndoRoBERTa untuk Klasifikasi Sentimen Opini Publik tentang Kecerdasan Buatan dalam Pendidikan di YouTube
DOI:
https://doi.org/10.30865/jurikom.v13i2.9697Keywords:
Sentiment Classification, Artificial Intelligence, YouTube, BERT, IndoRoBERTa, TransformerAbstract
This study aims to analyze the sentiment of Indonesian-language YouTube comments related to artificial intelligence (AI) in the field of education using a Transformer-based deep learning approach, namely Bidirectional Encoder Representations from Transformers (BERT) and IndoRoBERTa. The research data were obtained through the YouTube Data API, consisting of 10,834 comments reflecting public opinion on the implementation of AI in education. The dataset was manually labeled into three sentiment categories: positive, neutral, and negative, followed by a preprocessing stage including case folding, text cleaning, text normalization, tokenization, and topic filtering. The experimental results show that in the baseline scenario without fine-tuning, both models achieved low performance with accuracy below 41%. However, after fine-tuning, a significant improvement was observed, where IndoRoBERTa achieved an accuracy of 91.54% with an F1-score of 0.9134, while BERT reached an accuracy of 84.63% with an F1-score of 0.8413. These results indicate that Transformer-based models adapted to specific datasets are capable of better capturing the contextual and linguistic characteristics of informal and unstructured Indonesian text. In addition, IndoRoBERTa demonstrates more stable performance in handling class imbalance compared to BERT. Overall, this study demonstrates that Transformer-based approaches are effective for sentiment analysis in social media and can be used to more accurately and comprehensively understand public perceptions of the implementation of artificial intelligence in education.
References
[1] H. S. J. Chew dan P. Achananuparp, “Perceptions and Needs of Artificial Intelligence in Health Care to Increase Adoption,” vol. 24, hal. 1–19, 2022, doi: 10.2196/32939.
[2] R. Sahar dan M. Munawaroh, Artificial intelligence in higher education with bibliometric and content analysis for future research agenda. Springer International Publishing, 2025. doi: 10.1007/s43621-025-01086-z.
[3] M. Jamil, H. Hadiyanto, dan R. Sanjaya, “Sentiment Analysis: Classifying Public Comments on YouTube in Disaster Management Simulation in Indonesia Using Naïve Bayes and Support Vector Machine,” Ingénierie des systèmes d Inf., vol. 29, no. 2, hal. 437–446, Apr 2024, doi: 10.18280/isi.290205.
[4] J. Kamali, M. F. Alpat, dan A. Bozkurt, “AI ethics as a complex and multifaceted challenge : decoding educators ’ AI ethics alignment through the lens of activity theory,” Int. J. Educ. Technol. High. Educ., 2024, doi: 10.1186/s41239-024-00496-9.
[5] H. Zhu, Y. Sun, dan J. Yang, “Towards responsible artificial intelligence in education: a systematic review on identifying and mitigating ethical risks,” Humanit. Soc. Sci. Commun., 2025, doi: https://doi.org/10.1057/s41599-025-05252-6.
[6] A. Bello, S. C. Ng, dan M. F. Leung, “A BERT Framework to Sentiment Analysis of Tweets,” Sensors, vol. 23, no. 1, 2023, doi: 10.3390/s23010506.
[7] G. Yumitro, R. Febriani, A. Roziqin, dan A. Indraningtyas, “Bibliometric analysis of international publication trends on social media and terrorism by using the Scopus database,” Front. inCommunication, 2021, doi: https://doi.org/10.3389/fcomm.2023.1140461.
[8] H. Shaheen, “Social Media Marketing: A Bibliometric Analysis from Scopus,” Futur. Bus. J., 2024, doi: 10.1186/s43093-025-00465-2.
[9] D. Sun dan Y. Li, “behavioral sciences Influence of Strategic Crisis Communication on Public Perceptions during Public Health Crises : Insights from YouTube Chinese Media,” Behav. Sci. (Basel)., 2024, doi: https://doi.org/10.3390/bs14020091.
[10] A. Raza et al., “An improved deep convolutional neural network-based YouTube video classification using textual features,” Heliyon, vol. 10, no. 16, hal. e35812, 2024, doi: 10.1016/j.heliyon.2024.e35812.
[11] D. A. Musleh et al., “Arabic Sentiment Analysis of YouTube Comments: NLP-Based Machine Learning Approaches for Content Evaluation,” Big Data Cogn. Comput., vol. 7, no. 3, 2023, doi: 10.3390/bdcc7030127.
[12] S. M. Islam et al., “Challenges and future in deep learning for sentiment analysis : a comprehensive review and a proposed novel hybrid approach,” vol. 57, no. 3. Springer Netherlands, 2024. doi: 10.1007/s10462-023-10651-9.
[13] E. A. El-dahshan, M. M. Bassiouni, A. Hagag, R. K. Chakrabortty, H. Loh, dan U. R. Acharya, “RESCOVIDTCNnet : A residual neural network-based framework for COVID-19 detection using TCN and EWT with chest X-ray images,” Expert Syst. Appl., vol. 204, no. April, hal. 117410, 2022, doi: 10.1016/j.eswa.2022.117410.
[14] Y. Mao, Q. Liu, dan Y. Zhang, “Sentiment analysis methods , applications , and challenges : A systematic literature review,” J. King Saud Univ. - Comput. Inf. Sci., vol. 36, no. 4, hal. 102048, 2024, doi: 10.1016/j.jksuci.2024.102048.
[15] K. L. Tan, C. P. Lee, dan K. M. Lim, “A Survey of Sentiment Analysis : Approaches, Datasets, and Future Research,” Appl. Sci., 2023, doi: https://doi.org/10.3390/app13074550.
[16] X. Gong, W. Ying, S. Zhong, dan S. Gong, “Text Sentiment Analysis Based on Transformer and Augmentation,” Orig. Res., vol. 13, no. May, 2022, doi: 10.3389/fpsyg.2022.906061.
[17] S. Tzimiris, S. Nikiforos, M. N. Nikiforos, K. L. Kermanidis, dan D. Mouratidis, “A Comparative Evaluation of Transformer-Based Language Models for Topic-Based Sentiment Analysis,” Electron., 2025, [Daring]. Tersedia pada: https://www.mdpi.com/2079-9292/14/15/2957
