Title: Predicting elections over Twitter: a campaign strategies of political parties using machine learning algorithms

Authors: G. Anuradha; Nageswara Rao Moparthi; Sridhar Namballa

Addresses: Department of Computer Science and Engineering, Velagapudi Ramakrishna Siddhartha Engineering College, Vijayawada-AP, India ' Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, AP, India ' Department of Computer Science, Govt. Degree College, Tuni-AP, India

Abstract: Twitter is a social networking web application built to find out what is happening around us and all over the world. For this study, we analysed one million tweets for applying sentiment analysis is helpful to analyse the information where opinions are either good or bad/negative, or neutral and highly structured, heterogeneous in some cases. In this article we used a few machine learning algorithms like naïve Bayes, support vector machine, decision tree and neural network for analysis of elections results. Twitter data is used to predict the outcome of election by collecting data and analysing sentiments about the candidates. The experiential results show that the winning party's electoral success is expressively associated with social media/Twitter. It is a main promotional tool for new and upcoming parties in the 2019 general elections. This article mainly focused on Karnataka state elections of 2018 month of May for examining the analytical power of Twitter.

Keywords: Twitter; sentiment; naïve Bayes; support vector machine; SVM; decision tree; neural network.

DOI: 10.1504/IJAIP.2022.122196

International Journal of Advanced Intelligence Paradigms, 2022 Vol.21 No.3/4, pp.305 - 320

Received: 18 Sep 2018
Accepted: 27 Jan 2019
Published online: 12 Apr 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article