Analisis Sentimen Ulasan Pengguna Aplikasi YouCut Pada Google Play Store Menggunakan Metode Logistic Regression Classifier

Abstract

The rapid development of digital technology encourages people to be more active in reviewing digital services, including mobile applications. User reviews play an important role in shaping an app's reputation and success in the market. However, the high volume of reviews makes it difficult to analyze manually. This research aims to automatically analyze the sentiment of YouCut app user reviews with a machine learning approach using the Logistic Regression algorithm. The dataset in the form of 1,000 user reviews in Indonesian language is taken automatically from Google Play Store and processed through the stages of data cleaning, text preprocessing, lexicon-based sentiment labeling, feature extraction with TF-IDF, and model training and evaluation. The analysis results show that the Logistic Regression model is able to classify sentiment with an accuracy of 89%. Visualization of sentiment distribution shows the dominance of positive sentiment by 62.2%. Analysis of dominant words through wordcloud revealed that users appreciated the ease of use and video editing results, but also voiced complaints about performance and ad interruptions. The findings provide strategic insights for developers to continuously improve the features and user experience of the YouCut app.