Journal Article

Application of Twitter sentiment analysis in election prediction: a case study of 2019 Indian general election

Sikka, Geeta, Sharma, Nonita, Chauhan, Priyavrat

2023

Overview

Everyone must have experienced a huge collection of political content on their social media account’s home page during election time. Most of the users are busy in liking, sharing, and commenting political posts on the social media platform at that time, and these user activities show their attitude or behaviour towards the electoral or the political party. This study has mined the collective behaviour of Twitter users towards the Indian General election 2019. This work performed weekly sentiment analysis of massive Twitter content related to electoral and political parties during election time using a lexicon-based sentiment analysis approach. Based on this empirical study, the aim is to find out the feasibility of election prediction through social media analysis in a developing country like India. Further, an explorative analysis has been performed on the collected data, which gives answers to some dominant research hypotheses formulated in this paper. This paper shows how public mood can be gauged from social media content during the election period and how it can be considered as a parameter to predict the election results along with other factors. In addition, results evaluation has been done based on mean absolute error by considering the vote share and seat share of competing parties and leaders in the election. The predicted result of this work has been compared with exit poll results from various news agencies and the actual election result. It was found that our result of election prediction is quite similar to the final election results.

• Publisher: Springer Nature B.V
• Subject: Analysis / Attitudes / Candidates / Case studies / Collective behavior / Data collection / Data mining / Developing countries / Digital media / Election results / Elections / Emotions / Empirical analysis / Feasibility / Hypotheses / LDCs / Liking / Mass media / News / News agencies / Political alliances / Political parties / Politics / Predictions / Sentiment analysis / Social media / Social networks / Time use