Hybrid RoBERTa-BERTopic Sentiment Analysis for Information Governance at PT Bank Central Asia Tbk

Pastika Divasaputra; Deden Sumirat Hidayat

doi:10.33365/jtk.v20i2.2012

Pastika Divasaputra Esa Unggul University https://orcid.org/0009-0001-6122-6770
Deden Sumirat Hidayat Esa Unggul University https://orcid.org/0000-0002-1847-8665

DOI: https://doi.org/10.33365/jtk.v20i2.2012

Abstrak

Consumer reviews on digital distribution platforms offer valuable insights for financial institutions to evaluate service quality and user satisfaction. This study implements an integrated natural language processing (NLP) framework combining the pre-trained indonesian-roberta-base-sentiment-classifier model and BERTopic to perform granular sentiment analysis and thematic topic modeling on mobile banking user feedback. A total of 3,138 validated customer reviews were harvested from the Google Play Store. The sentiment classification revealed that 52.1% of the reviews expressed positive sentiments, while 36.0% were categorized as negative and 11.9% were categorized as neutral. Through cross-matrix BERTopic modeling, these sentiments were successfully mapped into critical operational themes, including application technical bugs, post-update system stability, and physical bank teller queue issues. The evaluation of the RoBERTa model demonstrated strong performance with an overall accuracy of 89.5% and a macro F1-score of 88.7%, proving its robustness in processing non-standard Indonesian text. The cross-analysis maps granular actionable insights, enabling banking management to systematically prioritize software engineering enhancements and customer service interventions based on quantified user pain points.

##submission.authorBiographies##

##submission.authorWithAffiliation##

Pastika Divasaputra is an undergraduate student in the Computer Science Department, majoring in Informatics Engineering at Universitas Esa Unggul, Jakarta, Indonesia. He currently serves as a Senior Banker at Bank Central Asia Tbk., where he is responsible for managing customer experience and satisfaction, as well as overseeing branch operations and business performance, including credit, lending, CASA, SME, cash management, investment, profitability, and overall financial outcomes. His academic and professional interests encompass information systems, business process optimization, data-driven decision-making, and financial technology applications. Pastika's scholarly profile and research contributions are registered at ORCID profile (https://orcid.org/0009-0001-6122-6770), and can be contacted via his institutional email at pdspro03@student.esaunggul.ac.id.

##submission.authorWithAffiliation##

Deden Sumirat Hidayat is a researcher at the National Research and Innovation Agency/ Badan Riset dan Inovasi Nasional (BRIN). He graduated as a doctor of computer science from the Faculty of Computer Science, University of Indonesia, Depok, Indonesia in July 2023. He obtained his master degree of Computer Sciences at Bogor Agricultural University in 2016. a He has an interest in the fields of KM, Information System, Management & Governance, data science, information and library science. He has experience in developing data mining-based knowledge management systems, particularly such as decision support systems. He also has experience as a web-based information system consultant and developer. Deden’s contributions to his field are accessible via his ORCID profile (https://orcid.org/0000-0002-1847-8665), and he is open to professional collaborations through dede025@brin.go.id.

Referensi

D. Chaffey and F. Ellis-Chadwick, Digital Marketing, 7th ed. Harlow, UK: Pearson Education, 2019.

PT Bank Central Asia Tbk, Sustainability Report 2024. Jakarta, Indonesia: PT Bank Central Asia Tbk, 2024.

M. Andrian, dkk., "Sentiment Analysis on Customer Satisfaction of Digital Banking in Indonesia," dalam Proc. International Conference on Information Management and Technology (ICIMTECH), Semarang, Indonesia, 2022, hlm. 436–441. doi: 10.1109/ICIMTECH55122.2022.9942213.

R. Adiningtyas and P. Auliani, "Sentiment Analysis for Mobile Banking Service Quality Measurement," 2024.

B. Liu, Sentiment Analysis: Mining Opinions, Sentiments, and Emotions. Cambridge, UK: Cambridge University Press, 2020.

B. Pang and L. Lee, "Opinion Mining and Sentiment Analysis," Foundations and Trends in Information Retrieval, vol. 2, no. 1–2, pp. 1–135, 2008.

Y. Liu, M. Ott, N. Goyal, et al., "RoBERTa: A Robustly Optimized BERT Pretraining Approach," arXiv:1907.11692, 2019.

A. Singgalen, "Performance Analysis of IndoBERT for Sentiment Classification in Indonesian Hotel Review Data," 2025.

N. Asri, et al., "Sentiment Analysis based on Indonesian Language Lexicon and IndoBERT on User Reviews PLN Mobile Application," 2025.

M. Grootendorst, "BERTopic: Neural Topic Modeling with a Class-Based TF-IDF Procedure," arXiv:2203.05794, 2022.

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, Minneapolis, MN, USA, 2019, pp. 4171–4186.

A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, "Attention Is All You Need," in Proc. Advances in Neural Information Processing Systems (NeurIPS), vol. 30, 2017.

B. Wilie, A. Vincentio, N. Winata, et al., "IndoNLU: Benchmark and Resources for Evaluating Indonesian Natural Language Understanding," arXiv:2009.05387, 2020.

R. Alawaji and A. Aloraini, "Sentiment Analysis of Digital Banking Reviews Using Machine Learning and Large Language Models," Electronics, vol. 14, no. 11, p. 2125, May 2025. doi: 10.3390/electronics14112125.

L. McInnes, J. Healy, and J. Melville, "UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction," arXiv:1802.03426, 2018.

R. J. G. B. Campello, D. Moulavi, A. Zimek, and J. Sander, "Hierarchical Density Estimates for Data Clustering, Visualization, and Outlier Detection," ACM Transactions on Knowledge Discovery from Data, vol. 10, no. 1, pp. 1–51, 2015.

V. Venkatesh, M. G. Morris, G. B. Davis, and F. D. Davis, "User Acceptance of Information Technology: Toward a Unified View," MIS Quarterly, vol. 27, no. 3, pp. 425–478, 2003.

F. D. Davis, "Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology," MIS Quarterly, vol. 13, no. 3, pp. 319–340, 1989.

A. Parasuraman, V. A. Zeithaml, and L. L. Berry, "SERVQUAL: A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality," Journal of Retailing, vol. 64, no. 1, pp. 12–40, 1988.

A. Parasuraman, V. A. Zeithaml, and A. Malhotra, "E-S-QUAL: A Multiple-Item Scale for Assessing Electronic Service Quality," Journal of Service Research, vol. 7, no. 3, pp. 213–233, 2005.

B. Pang, L. Lee, and S. Vaithyanathan, "Thumbs Up? Sentiment Classification Using Machine Learning Techniques," in Proc. Conference on Empirical Methods in Natural Language Processing (EMNLP), Philadelphia, PA, USA, 2002, pp. 79–86.

W. Medhat, A. Hassan, and H. Korashy, "Sentiment Analysis Algorithms and Applications: A Survey," Ain Shams Engineering Journal, vol. 5, no. 4, pp. 1093–1113, 2014.

Z. Yang, Z. Dai, Y. Yang, J. Carbonell, R. Salakhutdinov, and Q. V. Le, "XLNet: Generalized Autoregressive Pretraining for Language Understanding," in Proc. Advances in Neural Information Processing Systems (NeurIPS), vol. 32, 2019.

Z. Lan, M. Chen, S. Goodman, K. Gimpel, P. Sharma, and R. Soricut, "ALBERT: A Lite BERT for Self-supervised Learning of Language Representations," arXiv:1909.11942, 2019.

D. Araci, "FinBERT: Financial Sentiment Analysis with Pre-trained Language Models," arXiv:1908.10063, 2019.