Sentiment Analysis of Twitter for Election Prediction

Abstract

Elections are considered to be the most important feature of a democracy. In the past few years, election analysis and predictions have become very important for political parties and news organizations. The influx of various social media platforms such as twitter, Facebook and YouTube have drawn a large number of people that share their ideological and political thoughts and hence, it’s become important to analyse them in a much more sophisticated manner. Various data mining algorithms have been used to extract tweets and perform sentiment analysis pertaining to a related topic. Sentiment analysis refers to the technique to identify positive, negative or neutral opinions from a text. Though the use of sentiment analysis we will analyse the sentiment score for the two main political parties of India. The paper will brief on various techniques that have been used for election predictions. Various results from different methods have been included in this paper along with precision, accuracy and validity of the final outcome. The main aim of this paper is to create a model for the better prediction that will help in the analysis of voting choices of users. To increase the validity of the final results, various refining techniques have been used so that only relevant tweets are analysed.