Bi-Partisan And Broader Political Spectrum Bias Detection using Transformer Based LLM Models

Abraham, Abin Binoy (2025) Bi-Partisan And Broader Political Spectrum Bias Detection using Transformer Based LLM Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (584kB) | Preview

Abstract

Political biasness is a very common but a concerning issue that affects the overall neutrality and unbiased reporting of facts by different media houses. The popularity of social media platforms has allowed individual users and politicians to post information that can be fake as well highly biased towards a particular political ideology. This entails a broader political polarization of common masses that can result in undermining the overall democratic discourse. To address this, the thesis focuses on implementing AI techniques to measure bi-partisan political biasness in politically charged statements as well as to measure the political biasness in a much larger spectrum by headlines of news articles published by various media houses. To implement a solution to this Transformer based LLM models like BERT, RoBERTa and DeBERTa is used for fine tuning on three datasets; LIAR, UK News and QBias datasets to generate sentence embeddings for downstream classification. This model is compared with the baseline sentence embedding models like BG-BASE-EN to evaluate the performance. The proposed methodology outperformed all the baseline models in bi-partisan political bias classification as well as classification on a broader political spectrum. The BERT fine-tuned model outperformed the baseline models for bi-partisan classification by achieving a macro F1-score of 94% which is 20% better than the baseline model.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Razzaq, Abdul UNSPECIFIED
Uncontrolled Keywords:	Political Bias Detection; LLMs; BERT; RoBERTa; DeBERTa; Sentence Embeddings; MLP
Subjects:	J Political Science > JA Political science (General) Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4150 Computer Network Resources > The Internet > World Wide Web > Websites > Online social networks T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Telecommunications > The Internet > World Wide Web > Websites > Online social networks
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	11 Aug 2026 14:28
Last Modified:	11 Aug 2026 14:28
URI:	https://norma.ncirl.ie/id/eprint/9491

Actions (login required)

View Item