Layanan penulisan ilmiah yang disediakan oleh Perpustakaan Universitas Gunadarma
SENTIMENT ANALYSIS ON INDONESIAN TWEETS ABOUT KABINET INDONESIA MAJU USING LEXICON-BASED AND NAÏVE BAYES CLASSIFIER METHOD
ABSTRAKSI :
Joko Widodo as the elected president of Indonesia for the 2019-2024 period announced the composition of a new cabinet under the name of Kabinet Indonesia Maju. The election of new Ministers in Kabinet Indonesia Maju received mixed responses by the Indonesian citizen. The citizen is free to express their opinions about the election of the new Minister, not only positive or neutral opinions but also negative ones. Twitter as one of the most used social media in Indonesia provides a way for surveying public emotions about events or products associated with it. To be able to find out the tendency of sentiments on Twitter user tweets, we need a system that can analyze Tweets by applying sentiment analysis. The classification techniques used in this sentiment analysis are the Lexicon-Based and the Naive Bayes Classifier method. In Lexicon-Based classification, preprocessed datasets will be classified as positive, negative and neutral sentiments. Naive Bayes Classifier used to classify tweets into 6 emotions. The sentiment analysis results are implemented in the form of tables, bar charts, histograms, pie charts and wordcloud. The results of sentiment classification using the Lexicon Based method gets positive sentiment tendencies from a total of 5.693 tweet. Classification results using the Naive Bayes Classifier method get joy emotional tendencies from a total of 5.693 tweet. Classification using Naive Bayes Classifier method gets an accuracy of 82.17%.