Implementation of the Naive Bayes Method in Sentiment Analysis of Spotify Application Reviews

Abstract

This study focuses on sentiment analysis of Spotify application reviews on Google Play Store using the Naive Bayes algorithm. As a leading music streaming platform, Spotify receives diverse user feedback that reflects their experiences, complaints, and satisfaction. Sentiment analysis aids in understanding user opinions, enhancing services, and innovating features. The research involves collecting user reviews via web scraping, followed by preprocessing steps such as text cleaning, tokenization, normalization, stopword removal, and stemming. The Term Frequency-Inverse Document Frequency (TF-IDF) method is employed to assign weights to words, highlighting their significance in reviews. The Naive Bayes algorithm categorizes sentiments into positive, negative, and neutral classes. Performance evaluation uses a confusion matrix to measure accuracy, precision, recall, and F1-score. Results indicate that Naive Bayes effectively classifies large volumes of unstructured data with high accuracy and efficiency. This study contributes practically by offering actionable insights to improve Spotify's services and theoretically by advancing sentiment analysis methodologies using machine learning. The findings highlight the algorithm's potential to understand user needs and address issues, reinforcing its value in text analytics for mobile applications.