Uncovering Political Narratives in Bihar Assembly Elections2025:A Latent Dirichlet Allocation Based Topic ModelingStudyofReddit Discourse
IFR Journal of Humanities, Social Science and Politics · 01.03.2026 · 3 (1) · 10-20
DOI: 10.70146/hsspv03i01.002 ↗Kurz gefasst
Fachartikel in der Zeitschrift IFR Journal of Humanities, Social Science and Politics, erschienen am 01.03.2026. Verfasst von Swati Arya und 2 weiteren Autor:innen.
Abstract (Originalsprache)
This study investigates the latent thematic structures underlying online political discourse duringtheBihar Assembly Elections 2025. Using a corpus of 166,533 Reddit comments collected fromJanuary 2024toJuly 2025, Latent Dirichlet Allocation (LDA) has been applied to uncover dominant political narratives andissue-based clusters. Topic coherence optimization has been performed to determine the optimal number of topics, resulting in a five-topic model with a maximum coherence score of 0.675. Intertopic distance mappingreveals distinct thematic separations between caste-based discourse, electoral polarization, governance debates, public grievance discussions, and Hindi-dominant conversational exchanges. Results demonstrate strongpolarization around major parties (BJP, RJD, JD(U)), identity politics, reservation policies, and governanceevaluation. The findings highlight Reddit as a reflective space for issue-driven and emotionally chargedelectoral engagement. This work contributes to computational political communication research by providingastructured LDA-based discourse framework applicable to Indian state level elections.
Zitieren
APA
Arya, S., Mittal, D. A., & Agarwal, D. A. (2026). Uncovering Political Narratives in Bihar Assembly Elections2025:A Latent Dirichlet Allocation Based Topic ModelingStudyofReddit Discourse. IFR Journal of Humanities, Social Science and Politics, 3(1), 10-20. https://doi.org/10.70146/hsspv03i01.002
Vancouver
Arya S, Mittal DA, Agarwal DA. Uncovering Political Narratives in Bihar Assembly Elections2025:A Latent Dirichlet Allocation Based Topic ModelingStudyofReddit Discourse. IFR Journal of Humanities, Social Science and Politics. 2026;3(1):10-20. doi:10.70146/hsspv03i01.002
BibTeX
@article{Arya2026,
author = {Arya, Swati and Mittal, Dr. Ankush and Agarwal, Dr. Amit},
title = {Uncovering Political Narratives in Bihar Assembly Elections2025:A Latent Dirichlet Allocation Based Topic ModelingStudyofReddit Discourse},
journal = {IFR Journal of Humanities, Social Science and Politics},
year = {2026},
volume = {3},
number = {1},
pages = {10--20},
doi = {10.70146/hsspv03i01.002}
}
EUVIMED