Abstract
Aim : This research aims to analyze user sentiment towards Indonesian video streaming application services on the Google Play Store and map consumers of interest into categories of Over-the-top (OTT) service aspects.

Methods: A sentiment analysis model was developed using a fine-tuned IndoBERT model in conjunction with aspect-based sentiment analysis (ABSA) techniques to identify specific aspects within user reviews and classify their sentiment. A dataset of 32,000 reviews was utilized for this study. The research methodology followed the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework. A crucial step in the modeling phase was category mapping, which involved identifying critical aspects of OTT services from the dataset. Word clouds were employed to visualize the most frequent terms associated with each aspect.

Result: The fine-tuned IndoBERT model achieved impressive performance, with an accuracy of 95%, precision of 94%, recall of 94%, and an F1-score of 94%. This indicates the model's effectiveness in classifying sentiment towards specific aspects of OTT services. The final stage of the CRISP-DM deployment process involved creating interactive dashboards using Tableau to visualize the sentiment analysis results.

Conclusion: This research successfully employed a fine-tuned IndoBERT model and aspect-based sentiment analysis techniques to analyze user sentiment toward Indonesian video streaming application services on the Google Play Store. The model deployment through interactive dashboards facilitated the visualization and interpretation of sentiment analysis results. By leveraging these insights, OTT service providers can make data-driven decisions to improve user satisfaction, optimize their services, and gain a competitive edge in the market.