Analisis Sentimen Terhadap Layanan Provider Telekomunikasi Telkomsel Di Twitter Dengan Metode Naïve Bayes


  • Nizam Haqqizar Tanri Abeng University
  • Tika Nur Larasyanti Tanri Abeng University


Sentiment Analysis, Telkomsel, RapidMiner, Naïve Bayes


Opinion Data is taken from the Twitter social network based on a query in Bahasa Indonesia. This study aims to determine public sentiment on certain objects delivered on Twitter in Bahasa Indonesia, making it very useful to do market research on public opinion. The results of the study reviewed were 151 tweets with negative, neutral, positive sentiment &. Results of the 151 data processing training concluded that sentiment resulted in negative sentiment classification of 51 tweets, neutral sentiment of 51 tweets, and positive sentiments as much as 49 tweets. The accuracy rate in the category determination is 70.21% and the micro average of 70.20% in sentiment determination has a 70.11% precision rate and 70.33% return rate