Sentiment Analysis of Society Towards the Royalty Cases of Ahmad Dhani and Once Mekel Using Naïve Bayes

Abstract

In the case of royalties, this has become a topic of conversation which has raised pros and cons among the public on social media, namely on the Twitter platform. Through this public opinion, the reader can find out how the community thinks regarding the problems and royal dispute between Ahmad Dhani and Once Mikel. Opinions or opinions from the public or netizens on Twitter social media can be classified into positive, negative or neutral opinions, so that they can require sentiment analysis. There are several stages in this study using the R programming language starting from data collection which was carried out by collecting tweet data on social media Twitter as many as 300 tweets related to the cases of royalty Ahmad Dhani and Once Mekel, then pre-processing was carried out to look for words words that often appear in tweets. The last step is to use the K-Means method as a process of sorting or grouping 3 word clusters with the words often, moderate and rarely used in tweets. By testing based on the word tweet obtained an accuracy rate of 62.6% while testing based on daily tweets obtained an accuracy rate of 65.3%.

Keywords: K-Means, Tweets, Royalties, Sentiment Analysis.