Klasifikasi Opini Masyarakat Terhadap Jasa ISP MyRepublic dengan Naïve Bayes

Hafiz Irsyad, Ahmad Farisi, Muhammad Rizky Pribadi
2019 Jurnal Nasional Teknik Elektro dan Teknologi Informasi (JNTETI)  
Opinion classification is an analysis that aims to determine the sentiments of the community or a group about a particular entity. Opinion classification can be categorized as positive, negative, and neutral. This research of the classification of public opinion was conducted on the MyRepublic internet service provider. At the moment, MyRepublic has reached seven provinces in Indonesia. MyRepublic has used a lot of media to communicate with its customers, especially Twitter. MyRepublic Twitter
more » ... MyRepublic Twitter account is MyRepublicid with a number of followers of 9,414. This research uses comments or tweets from followers that can be used to see opinions from followers of My Republic, whether positive or negative. The comments or tweets classification on Twitter is using naïve Bayes method. The data used is 1,553. As much as 70% of the data from each category is used as training data and the remaining 30% as testing data. The naïve Bayes method produces positive accuracy value of 0.976%, negative accuracy value of 0.82895%, and neutral accuracy value of 0.8333%, with an average of 0.87949%. Based on the result, it can be concluded that the naïve Bayes method is able to classify the data very well.
doi:10.22146/jnteti.v8i1.487 fatcat:lpepcmp6trf7blsmsikd2w7sku