Shashvat M's Senior Project Blog

Project Title:  Framing Bias: An AI-Powered Analysis of Political Media Narratives BASIS Advisor: Mr. Kwit Internship Location: Remote Onsite Mentor: Sibish Basheer

Project Abstract

Media bias significantly influences public opinion and political discourse, often deepening ideological divisions. My research examines political biases in news media using machine learning and large language models (LLMs). By systematically investigating patterns in news coverage, this project highlights how various news outlets frame political events and narratives. It explores the extent of political bias among major news sources and evaluates how effectively machine learning and LLMs can identify and quantify this bias. To conduct this research, I compiled a dataset of news articles from multiple sources that focused on political topics over a defined period. Employing natural language processing (NLP) techniques, sentiment analysis, and supervised machine learning classifiers, I trained models to recognize patterns of political bias. I also leveraged LLMs to analyze contextual framing, emotional tone, and linguistic cues indicative of partisan leanings. The findings have significant implications for media literacy, journalism, and AI-based fact-checking. By providing a data-driven approach to analyzing bias, this research promotes more balanced news consumption and aids in developing tools for readers to critically evaluate media reports.