Strategy For Increasing the Brand Reputation of The PLN Mobile Application Based on Social Media Sentiment Analysis Using Machine Learning

Abstract

Social media is currently a trendy medium for people in Indonesia to express opinions. Users can easily express their experiences with a product through social media, including the PLN Mobile application from PT PLN (Persero). The application becomes a digital platform to meet various customer needs related to electricity services. Twitter social media has provided data on all tweets and reviews that can be accessed with certain keywords publicly through the Twitter API (Application Programming Interface). This research will use the text mining method with the word cloud approach, network explorer, types of emotions, and sentiment analysis, which can be used to analyze user opinions. The tools used are Orange Data Mining, which applies text preprocessing, including transformation, tokenization, normalization, and filtering, which aims to analyze text. The method used for sentiment analysis classification of Twitter users' opinions is VADER: Lexicon- and Rule-Based Sentiment Analysis. This method analyzes and classifies sentiment on social media towards the PLN Mobile application in Indonesia. The resulting classification model can be used as an early warning about how belief occurs on social media towards the PLN Mobile application. In addition, several recommendations on managerial aspects were also produced referring to the research data.