Setiment Analysis of Public Opinion on the Go-Jek Indonesia Through Twitter Using Algorithm Support Vector Machine

Abstract

The development of technology and information, especially in Indonesia is very rapid so that social media is the most popular communication tool by the people of Indonesia today. One of these social media is Twitter. This also causes the public to tend to give opinions and assessments in the form of tweets to service companies, one of which is Go-Jek Indonesia. Public opinion and judgment on Twitter can be classified into 3 classes: negative, neutral, and positive. The purpose of this study is to analyze the sentiment of public opinion on Go-jek Indonesia on twitter using the Support Vector Machine (SVM) algorithm. The approach used were Multiclass One Vs Rest SVM with Univariate Chi Square feature selection to classify community tweets on Go-Jek Indonesia's services. Using testing data of 170 tweets, 31.2% of people with negative opinions were obtained, 24.1% were neutral and 36.5% were positive opinions and 5.9% failed to be classified. The results of sentiment analysis testing conducted provide a classification accuracy of 91.8%.