Ulasan Pengguna Sebagai Cerminan Kualitas Media Digital: Studi Analisis Sentimen Pada Aplikasi Detikcom

Abstract

This study aims to analyze user review sentiments as a reflection of digital media quality in the Detikcom news application on Google Play Store. User reviews are considered a form of evaluative communication that represents audience perceptions and experiences toward digital media. This research employs a quantitative approach using text-based sentiment analysis. The data consist of 2,025 user reviews collected from Google Play Store and analyzed using Python programming. The analysis stages include text preprocessing, sentiment classification into positive, neutral, and negative categories, word frequency analysis, and examination of the relationship between sentiment and user ratings using the Chi-square test. The results indicate that positive sentiment dominates user reviews, suggesting a generally favorable perception of content quality and news update speed. Nevertheless, a considerable proportion of negative sentiment is identified, mainly related to excessive advertising and technical issues within the application. The Chi-square test reveals a statistically significant relationship between numerical ratings and user review sentiment (χ² = 4002.00; p < 0.001), confirming that ratings correspond with users’ emotional expressions in textual reviews. This study concludes that sentiment analysis of user reviews serves as a valuable indicator for assessing the quality and credibility of digital media and contributes to the development of audience-centered communication studies in the digital media context.