Perbandingan Kinerja Naïve Bayes, Support Vector Machine, dan K-Nearest Neighbor dalam Analisis Sentimen Mobile Legends

Abstract

The rapid advancement of information and communication technology has significantly increased the popularity of online games in Indonesia, one of which is Mobile Legends: Bang Bang (MLBB), with millions of active users. The abundance of user reviews on digital platforms provides valuable data for analysis using text mining and natural language processing (NLP) approaches. Sentiment analysis is applied to classify user opinions into positive, negative, and neutral categories, offering insights into player satisfaction and perceptions of game quality. This study compares the performance of three classification algorithms, Naïve Bayes (NB), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN), in analyzing sentiment from Mobile Legends user reviews on the Google Play Store. A total of 5,000 reviews were collected using the web scraping technique and processed through the Knowledge Discovery in Databases (KDD) framework, which includes cleaning, case folding, tokenization, normalization, and stopword removal. Sentiment labeling was performed using a lexicon-based approach with the InSet sentiment lexicon. The dataset was divided into training and testing sets with an 80:20 ratio and evaluated using accuracy, precision, recall, and F1-score metrics. The results indicate that the SVM algorithm achieved the highest accuracy of 88.1%, followed by KNN at 65.1% and NB at 62.6%. Thus, SVM is recommended as the most effective model for sentiment analysis of Mobile Legends user reviews.