SENTIMENT ANALYSIS OF ONLINE GOJEK TRANSPORTATION SERVICES ON TWITTER USING THE NAÏVE BAYES METHOD

Abstract— Social media is the most accessed internet content by internet users in Indonesia. This is not surprising, given the many benefits that social media provides, one of which is the benefit of self-expression. Self-expression can include many things, including emotional openness, which is the openness of a person in conveying the emotions he is feeling. Along with the development of social media, emotional disclosure is ubiquitous in social media, one of which is social media Twitter. With the development of information technology, means of transportation are also developing with the existence of online transportation services. Currently, the use of online transportation services has become a necessity, so it is necessary to conduct a sentiment analysis on online transportation services to find out how the public responds to these online transportation services. The purpose of this study is to analyze community responses by analyzing data in the form of tweets and then classifying them into positive, negative, and neutral classes using the Naïve Bayes method because the error rate obtained is lower when the dataset is large, besides that the accuracy of Naive Bayes and the speed is higher. high when applied to a larger dataset. The results of this study indicate that the neutral sentiment level of public tweets is greater than the level of positive sentiment and negative sentiment with an accuracy of 25.00%.