Sentiment Analysis on the Level of Customer Satisfaction to Data Cellular Services Using the Naive Bayes Classifier Algorithm

Abstract

Internet users in Indonesia have increased in recent years. Many product service providers who provide internet access services in accordance with tariff options and their superiority. In this research, sentiment analysis on social media to some service data service operator to see the level of public satisfaction in using data service of telecommunication operator for internet access in Indonesia.In this research is sentiment analysis with several stages, namely the collection of sentiment data using API (Application Programming Interface) which is available on Twitter. The preprocessing stage is then processed to process raw initial data, then perform POS tagging and weighing the word with TF-IDF calculation and perform classification using the Naive Bayes Classifier (NBC) method. This study yields an average value of 94,5% precision rate, 93,3% Recall and 99,09% Accuracy.