Title: Analysing political opinions using machine learning

Authors: Pragya Joshi; Akash Singh Kunwar

Addresses: Cognizant, DLF Phase 3, Sector 24, Gurugram, Haryana, India ' Simon Fraser University, Burnaby, BC, Canada

Abstract: In the era of digital world, text is not confined to textbooks or newspapers anymore. People use platforms like Twitter, Facebook, Quora, and other social media platforms to express their opinions over certain products, movies, social, economic or political causes. Huge chunks of textual data are available on these platforms for analysis. This paper tries to leverage deep learning and natural language processing (NLP) to use the publicly available text data to predict outcomes of Indian general elections by analysing the tweets with hashtags for various parties, using opinion mining to define polarity in the opinions. It tries to adopt a hybrid approach using NLP. The results from the analysis help in highlighting the potential of machine learning in predicting the election results and identifying the political inclination of people towards specific policies thus, indicating the efficiency of using social media to predict real-world outcomes.

Keywords: Twitter; sentiment analysis; machine learning; decision tree; random forest.

DOI: 10.1504/IJPSPM.2023.135035

International Journal of Public Sector Performance Management, 2023 Vol.12 No.4, pp.524 - 539

Received: 21 Mar 2020
Accepted: 01 Jul 2020
Published online: 28 Nov 2023 *

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