Surabaya Government Performance Evaluation Using Tweet Analysis

Abstract

The purpose of this research is to determine the various positive attributes appreciated by the public, and the negative things that need to be improved by the Surabaya government. The sentiment analysis methods, including the Naïve Bayes Classifier, Support Vector Machine, and Logistic Regression, are employed to classify the pros and cons of the Surabaya government. The comparison of the three methods demonstrated that SVM gives the best classification accuracy compared to others. Police performance is the highlighted word in the positive category, while traffic congestion is in the negative.