Sentiment analysis of public opinions on the welfare of honorary educators using Naive Bayes

Abstract The development of the internet in Indonesia is very rapid, this is marked by the number of emerging social media. One of the most popular social media is Twitter. Many Twitter users who tweet their expression and opinion on a product, service, political issues or things that are trending. Even the current government is not spared from public comment on social media, one of which is an honorary teacher at an elementary school in Pandeglang Regency, who lives next to his school toilet. This case was widely discussed by almost all Indonesians through offline and online media, especially on social media. The impact of this case is the difference in response in the form of public sentiment towards public services, especially in improving the welfare of educators. Various sentiments of Netizens (internet users) appear on social media ranging from praise, criticism, suggestions, satire and even expressions of hatred towards the Pandeglang Government. This research was conducted to determine netizens’ sentiments towards the Pandeglang Regency government based on the case of honorary teacher news in Pandeglang Regency who live in school toilets. The approach used is Naive Bayes. The results of sentiment analysis showed that there are 16 Twitter posts grouped in neutral class, 19 posts in positive class and the remaining 16 posts are in the negative class. The level of accuracy generated from the sentiment analysis model formed is 88.24%.