ANALISIS SENTIMEN PESERTA MAGANG STUDI INDEPENDEN BERSERTIFIKAT (MSIB) KEMENDIKBUD PADA TWITTER MENGGUNAKAN NAÏVE BAYES

The problem faced by the apprentices is that there are complaints and counter-opinions about the MSIB program which are spread through social media. Therefore, a method is needed to analyze sentiment and help apprentices identify the authenticity of information. The method used is sentiment analysis using the Naive Bayes Classifier, which has been proven to have a fairly high accuracy in processing data. In this study, the model built reached an accuracy level of 0.7794. In addition, the model performance evaluation shows a precision of 0.7823, a recall of 0.7794, and an F1-score of 0.7805. The purpose of this thesis is to measure the level of satisfaction of MSIB Kemendikbud apprentices through sentiment analysis on the Twitter platform. Researchers hope that the results of the sentiment analysis can help apprentices identify correct information and improve the quality of learning during the internship program.

Keyword: MSB, Twitter, Naïve Bayes