SOCIAL MEDIA DISCOURSE AND SENTIMENT TOWARD U.S. PRESIDENTIALCANDIDATES:ATEXTANALYSISOFTWITTERPOSTSMENTIONINGJOE BIDEN ANDDONALD TRUMP

Abstract

Socialmedia has emerged as a significant platform of political communication where the citizenryairstheir viewsand argueswith each other regarding political leaders. Thisstudyhas analyzed social media discussions and theattitude topresidential candidates of the United States of America using text analysis of Twitter accounts involving thementioningof Joe Biden and Donald Trump. The data was composed of 5328 tweets and was taken publicly andcomprised of 2540 tweets mentioning the name Biden and 2788 tweets mentioning the name Trump. There search used quantitative text analysis on the language and mood that was conveyed and expressed in these tweets. Sentiment analysis was the classification of Twitter users who posted positive, negative, and neutral messages, andkeyword frequency analysis wasa technique to determine the prevailing discourse patterns. The findings revealed that thepercentage of negative sentimentof tweets was the biggest proportion of tweets concerning the two candidates, and theproportions of neutral and positivesentimentwerelower.Referredwordslikepresident,people,and Americawereusedrepeatedly,whichindicatedthatpeople often expressed their remarks in greater political and national grounds.The slight differences in the distributionof sentiments between the tweets that mentioned the two candidates were alsoidentified when comparing them. The resultsshow that Twitter is an active political discourse where people express
their views and emotions about political leaders.The work can be of use to research in the field of media and communicationby demonstrating that social media data can be analyzed through text analysis to identify the trends of the politicaldiscourse and sentiment in modern digital spaces.