Political Pulse of India: AI-Based Multilingual Analysis of Social Media Discourse and Election Trends

PoliPulse is a platform that uses intelligence to analyze politics. It looks at what happened in elections and what people are saying about politics on social media right now. The platform uses computer programs to understand what people are saying and how they feel about politics. The system collects information from places like Twitter, Reddit, news websites and election databases. This helps us see how people feel about politics, which parties they support what they are talking about and how their opinions change over time. The platform uses some techniques like RoBERTa and BERTopic to understand what people mean and how they feel. It also uses a kind of computer program called Random Forest to make predictions. When we tested PoliPulse it worked well with information in many languages. It could handle a lot of information. Still give us answers quickly. The part of the platform that figures out how people feel was correct about 88.5% of the time. The whole system worked well. We also made a website that shows us the information in a way that is easy to understand. It has graphs and pictures that help us see what people are talking about and how they feel. PoliPulse solves some problems that come up when we try to understand what people are saying about politics in many languages. It can handle information from sources understand complex ideas and process a lot of information quickly. By using intelligence and real-time information PoliPulse gives us a practical way to analyze politics. Overall PoliPulse shows us how artificial intelligence can help us understand politics better. It brings together computer science, artificial intelligence and real-time information to help us see what people think about politics. This can help researchers, politicians, analysts, journalists and institutions understand what people think and how they feel about politics. It can also help us make better systems to analyze elections in the future. PoliPulse is a tool, for understanding politics and PoliPulse can help us make sense of it all.

[1]     M. G. Folgado and V. Sanz, “Exploring the political pulse of a country using data science tools,” 2022. [2]     A. Tumasjan, T. O. Sprenger, P. G. Sandner, and I. M. Welpe, “Predicting elections with Twitter: What 140 characters reveal about political sentiment,” in Proceedings of the International AAAI Conference on Web and Social Media (ICWSM), 2010, pp. 178–185. [3]     J. Pal, P. Chandra, and V. G. Vydiswaran, “Twitter and the rebranding of Narendra Modi,” Economic & Political Weekly, vol. 51, no. 8, pp. 52–60, 2016. [4]     D. Kakwani et al., “IndicNLPSuite: Monolingual corpora and pretrained multilingual models for Indian languages,” in Findings of ACL: EMNLP, 2020, pp. 4948–4961. [5]     F. Pierri, C. Piccardi, and S. Ceri, “Twitter as a proxy for political programmes, parties, and leaders: Measuring the political discourse using data science tools,” Social Network Analysis and Mining, vol. 11, no. 157, pp. 1–14, 2021. doi: 10.1007/s42001-021-00157-1. [6]     A. Kumar and T. Sebastian, “Sentiment analysis of Twitter data during Indian elections,” Journal of Information and Computational Science, vol. 11, no. 9, pp. 453–463, 2021. [7]     A. Joshi, A. R. Balamurali, P. Bhattacharyya, and M. Carman, “Towards sub-word level compositions for sentiment analysis of Hindi-English code-mixed text,” in Proceedings of COLING, 2016, pp. 1310–1319. [8]     V. Gupta, A. Joshi, and P. Kumaraguru, “Deep learning-based sentiment analysis for Indian code-mixed social media text,” in Proceedings of ACM India CoDS-COMAD, 2019, pp. 65–73.