Augmentasi GPT-4o dan Fine-Tuning Indobert untuk Analisis Sentimen Publik Pada Isu Reshuffle Menteri Keuangan
Abstract
Reshuffle Menteri Keuangan pada tahun 2025 memicu perhatian publik luas dan menghasilkan dinamika opini di media sosial, khususnya X. Opini publik yang terekam dalam bentuk teks bersifat masif, real-time, dan tidak terstruktur, sehingga menghadirkan tantangan analisis karena penggunaan bahasa informal serta distribusi kelas sentimen yang tidak seimbang. Penelitian ini bertujuan untuk mengidentifikasi sentimen publik terkait reshuffle Menteri Keuangan dengan memanfaatkan integrasi GPT-4o dan IndoBERT. GPT-4o digunakan sebagai instrumen augmentasi data untuk memperkaya kelas minoritas, sedangkan IndoBERT berperan sebagai model klasifikasi sentimen yang dioptimalkan untuk bahasa Indonesia. Hasil penelitian menunjukkan bahwa pendekatan ini mampu meningkatkan kualitas representasi data dan stabilitas klasifikasi. Model IndoBERT yang dilatih dengan data hasil augmentasi mencapai akurasi 86% dan macro-F1 sebesar 0,86, dengan performa terbaik pada kelas negatif (F1=0,88), disusul positif (F1=0,86) dan netral (F1=0,83). Temuan ini menegaskan bahwa integrasi GPT-4o dan IndoBERT efektif dalam mengatasi imbalanced data serta meningkatkan keandalan analisis sentimen berbahasa Indonesia. Penelitian ini tidak hanya memperkaya literatur analisis teks di Indonesia, tetapi juga memberikan kontribusi praktis bagi pembuat kebijakan dan media dalam memahami opini publik secara lebih proporsional.
Full text article
References
Alkhawaldeh, I. M., Albalkhi, I., & Naswhan, A. J. (2023). Challenges and limitations of synthetic minority oversampling techniques in machine learning. World Journal of Methodology, 13(5), 373–378. https://doi.org/10.5662/wjm.v13.i5.373
Aras, S., Yusuf, M., Ruimassa, R. Y., Wambrauw, E. A. B., & Pala’langan, E. B. (2024). Sentiment Analysis on Shopee Product Reviews Using IndoBERT. Journal of Information Systems and Informatics, 6(3), 1616–1627. https://doi.org/10.51519/journalisi.v6i3.814
Ashar, M. N., & Siahaan, D. O. (2024). Text Augmentation to Overcome Data Limitations in Sentiment Analysis for Bahasa Indonesia. Proceedings of 2024 IEEE International Conference on Data and Software Engineering: Data-Driven Innovation: Transforming Industries and Societies, ICoDSE 2024, 217–222. https://doi.org/10.1109/ICODSE63307.2024.10829895
Baharuddin, F., & Naufal, M. F. (2023). Fine-Tuning IndoBERT for Indonesian Exam Question Classification Based on Bloom’s Taxonomy. Journal of Information Systems Engineering and Business Intelligence, 9(2), 253–263. https://doi.org/10.20473/jisebi.9.2.253-263
Canesta, F., & Roestam, R. (2024). Influencer pricing prognostication on social media dynamics: An advanced examination of the linear regression polynomial degree algorithm & neural networks. MUST: Journal of Mathematics Education, Science and Technology, 9(2), 76–91. https://doi.org/10.30651/must.v5i1.21040
Carbonell, J. (1998). The Use of MMR, Diversity-Based Reranking for Reordering Documents and Producing Summaries.
Carvalho, M., Pinho, A. J., & Brás, S. (2025). Resampling approaches to handle class imbalance: a review from a data perspective. Journal of Big Data, 12(1). https://doi.org/10.1186/s40537-025-01119-4
Duffy, W., O’Connell, E., McCarroll, N., Sloan, K., Curran, K., McNamee, E., Clist, A., & Brammer, A. (2025). Evaluating rule-based and generative data augmentation techniques for legal document classification. Knowledge and Information Systems. https://doi.org/10.1007/s10115-025-02454-x
Jim, J. R., Talukder, M. A. R., Malakar, P., Kabir, M. M., Nur, K., & Mridha, M. F. (2024). Recent advancements and challenges of NLP-based sentiment analysis: A state-of-the-art review. Natural Language Processing Journal, 6, 100059. https://doi.org/10.1016/j.nlp.2024.100059
Lin, C. H., & Nuha, U. (2023). Sentiment analysis of Indonesian datasets based on a hybrid deep-learning strategy. Journal of Big Data, 10(1). https://doi.org/10.1186/s40537-023-00782-9
Papineni, K., Roukos, S., Ward, T., & Zhu, W.-J. (2001). BLEU. Proceedings of the 40th Annual Meeting on Association for Computational Linguistics - ACL ’02, 311. https://doi.org/10.3115/1073083.1073135
Rodríguez-Ibánez, M., Casánez-Ventura, A., Castejón-Mateos, F., & Cuenca-Jiménez, P. M. (2023). A review on sentiment analysis from social media platforms. In Expert Systems with Applications (Vol. 223). Elsevier Ltd. https://doi.org/10.1016/j.eswa.2023.119862
Setyo Nugroho, K., Yullian Sukmadewa, A., Wuswilahaken, H. D., Abdurrachman Bachtiar, F., & Yudistira, N. (2021). BERT Fine-Tuning for Sentiment Analysis on Indonesian Mobile Apps Reviews. https://research.google/teams/brain.
Sokolova, M., & Lapalme, G. (2009). A systematic analysis of performance measures for classification tasks. Information Processing and Management, 45(4), 427–437. https://doi.org/10.1016/j.ipm.2009.03.002
Suhaeni, C., & Yong, H. S. (2023). Mitigating Class Imbalance in Sentiment Analysis through GPT-3-Generated Synthetic Sentences. Applied Sciences (Switzerland), 13(17). https://doi.org/10.3390/app13179766
Wei, J., & Zou, K. (2019). EDA: EASY DATA AUGMENTATION TECHNIQUES FOR BOOSTING PERFORMANCE ON TEXT CLASSIFICATION TASKS.
Wozniak, S., & Kocon, J. (2023). From Big to Small Without Losing It All: Text Augmentation with ChatGPT for Efficient Sentiment Analysis. IEEE International Conference on Data Mining Workshops, ICDMW, 799–808. https://doi.org/10.1109/ICDMW60847.2023.00108
Authors
Copyright (c) 2025 Sabrina Adnin Kamila, Aisya Wina Wahda, Baiq Nina Febriati, Anwar Fitrianto, Rachmat Bintang Yudhianto
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-NonCommercial 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work