Twitter Sentiment Analysis of Indonesia Internet Service Provider: A Case Study of Indihome and Firstmedia

Abstract

Every company, especially internet service providers, needs to improve the quality of its products. The initial stage of improving product quality can be done by knowing customer satisfaction as an evaluation material. One way to determine customer satisfaction with a product is by calculating Customer Satisfaction using sentiment analysis on Twitter data, where Twitter is a very popular social media to get customer satisfaction feedback. The sentiment analysis method used is machine learning using a decision tree and random forest. From the two methods, the results show that the random forest method is better than the decision tree method. The oversampling method is also carried out to overcome the imbalanced class data set. The results of the Customer Satisfaction Score obtained show that tweets containing the word "Firstmedia" are better than tweets containing the word "Indihome" with a value of -41.67 for Firstmedia and -59.259 for IndiHome