Optimization of Sentiment Analysis for Indonesia Presidential Election using Naïve Bayes and Particle Swam Optimization
Nur Hayatin, Gita Indah Marthasari, Lia Nuraini
Jl. Raya Tlogomas 246, Malang, Indonesia
Department of Informatic Engineering, Faculty of Engineering.
Universitas Muhammadiyah Malang
Abstract
Knowing peoples sentiments on social media, especially from Twitter, is very interesting to analyze. especially the analysis of sentiments related to presidential candidates in the 2019 election in Indonesia. This study aims to extract opinions from twitter feeds to find out the results of public sentiment in Indonesia Elections. This research using naïve Bayes method with PSO to classification twitter feeds. PSO is used in the feature selection process to find optimization values to improve the accuracy of Naïve Bayes. There are 3 main stages of the process, i.e. preprocessing, feature extracting, and classifying. From this study, the group of tweets was obtained based on the positive and negative sentiments from community towards two presidential candidates of Indonesia in 2019. The testing result shown the accuracy 90.74% with optimization using Naive Bayes with PSO.
Keywords: sentiment analysis; opinion extraction; naive bayes; social media; classification
Topic: International Symposium of Engineering, Technology, and Health Sciences