Abstract: Use of Sentiment analysis is increasing day by day know the sentiments of public on social media. It analyzes sentiments, emotions and opinion of people which are shared on social sites. Twitter is a powerful social networking tool used frequently to online sharing of views. In this paper, tidytext mining technique is used for sentiment classification of twitter data. Sentiment analysis is performed on tweets of two popular Indian leaders' for prediction of 2019 election which are extracting from twitter posts from 2014 to 2018, and for this R language is utilized that provides Tidy text package. Tidytext package contains three lexicons NRC, BING, and AFINN to analyze emotions and comparison of three lexicons.

Keywords: Twitter, sentiment analysis, Tidytext lexicon approach: NRC, BING, AFINN, and wordcloud, Tidy text mining

| DOI: 10.17148/IARJSET.2020.7110