Sentiment Analysis on User Satisfaction Level of Cellular Data Service Using the K-Nearest Neighbor (K-NN) Algorithm

In today's modern era, social media is very close to people's lives. For each person can have up to more than 2 accounts for each social media such as Twitter, Instagram, Facebook, LINE, Path, and so forth. This makes the social media as the largest data collection of opinion from the public or internet users. To be able to retrieve data and draw conclusions of positive and negative values of an opinion on social media then do analysis of sentiment. The author analyzed the sentiments on the satisfaction of the telecommunication operator service users to the telecommunication service provider in Indonesia from each of their own official accounts or by using keywords related to telecommunication service providers in Indonesia. In performing the analysis, the author will use K-Nearest Neighbor (K-NN) analysis method with TF-IDF and Part-of-Speech (POS) Tagging. The results of this study obtained the average value of Precision 92,21 %, Recall 93,74%, F1-score 92,20%, and Accuracy 98,94%.