ANALISIS SENTIMEN DONALD TRUMP DAN JOE BIDEN PADA PENGGUNA TWITTER SEBELUM PEMILIHAN PRESIDEN 2020 DENGAN LEXICON BASED

The use of social media for communication has increased rapidly in recent years. The important role of social media, especially Twitter, has been proven to influence Donald Trump's victory in the 2016 American election. A significant aspect of Twitter is how Twitter can be involved in the realm of world politics. The purpose of the study was to compare the differences in sentiments of Twitter users' reactions in tweeting against Donald Trump and Joe Biden. The method used in this research is lexicon based. The lexical words used are 6800 negative and positive words with a sentiment classification of -1 for negative, 0 for neutral, and 1 for positive. The results of the classification process carried out using 90,000 tweet data related to Donald Trump and Joe Biden with the lexicon based method obtained sentiments related to Donald Trump as follows: 38% for negative sentiment, 45% for neutral sentiment, and 16% for positive sentiment. Meanwhile, for Joe Biden, negative sentiment was 35%, 45% for neutral sentiment, and 19% for positive sentiment.