Analisis Sentimen Game Mobile Legends Berdasarkan Review Pengguna di Playstore Menggunakan Algoritma Naive Bayes

Abstract

Sentiment analysis of Mobile Legends user reviews on the Google Play Store was conducted using the Naive Bayes algorithm, with the aim of identifying user perceptions and providing recommendations for developers to improve the quality of the game. The methods used include data collection, text processing, and vectorization using TF-IDF. The data was divided into training and testing subsets to build and evaluate the model. Results showed an accuracy of 74%, with precision 1.00 and recall 0.00 for positive sentiment, and precision 0.74 and recall 1.00 for negative sentiment. Although the model effectively detects negative sentiment, optimization is needed to improve the detection of positive sentiment. The findings provide valuable insights for developers in understanding user opinions and improving game quality.