An experimental study of lexicon-based sentiment analysis on Bahasa Indonesia

We can get various information only by accessing a certain website. Almost every task can be done by using digital technology. Internet users that increase rapidly also bring more useful data that we can make further analysis. One of the data that we can use is user opinion data. This information helps to make a decision, market research, and search engine. A technique that we can use to obtain this information is opinion mining or usually called sentiment analysis. There are two methods available, supervised and lexicon-based sentiment analysis. The supervised method has better performance than the lexicon-based method. However, supervised method performance is very dependent on the quality and the amount of training data. In this study, we would try to implement lexicon-based sentiment analysis on Indonesian data opinion. Overall, our approach has an accuracy of 0.68. This result is quite good as starting point for further research. This study is expected to trigger another study in lexicon-based sentiment analysis, especially for Bahasa Indonesia.