Comparison of the Naïve Bayes Algorithm and Support Vector Machine in Sentiment Analysis of "Teman Bus" Application

Perbandingan Algoritma Naive Bayes dan Support Vector Machine dalam Analisis Sentimen Aplikasi teman Bus

Abstract

This research compares the performance of two classification algorithms, Naïve Bayes and Support Vector Machine, in analyzing sentiment from user reviews of the Teman Bus application on Google Play Store. While both algorithms have been used in sentiment analysis before, direct comparison within the context of the Teman Bus application has not been done. The sentiment analysis process involves text processing and sentiment classification to evaluate user responses to the bus service.

The comparison results show variation in sentiment classification performance between Naïve Bayes and Support Vector Machine. Support Vector Machine, particularly with the RBF kernel, demonstrates an accuracy of 85%, excelling in handling complex and non-linear sentiment patterns compared to Naïve Bayes, which has an accuracy of 82%, especially on datasets with a 30:70 data split ratio.

This research provides deeper insights into user evaluations of the Teman Bus service through reviews on Google Play Store and offers valuable insights into selecting the appropriate algorithm for similar tasks in the future. Support Vector Machine with the RBF kernel tends to be a preferred choice for analyzing sentiment in the Teman Bus application; however, the choice of the best algorithm depends on the context and characteristics of the data.