The Media Bias Detector: A framework for annotating and analyzing the news

Abstract

News organizations introduce bias into their coverage via the choices they make about which topics to cover (or ignore) and how to frame the issues they do decide to cover. Here, we introduce the Media Bias Detector, a scalable computational framework that integrates large language models (LLMs) with near-real-time news scraping to extract structured annotations, including political lean, tone, topics, article type, and major events, across hundreds of articles per day. We quantify these dimensions of coverage at the sentence level, the article level, and the publisher level, expanding the ways in which researchers can analyze selection and framing bias in the modern news landscape. We also release an interactive web platform for convenient exploration of these data and an accompanying dataset covering more than 140,000 articles published in 2024 by 10 prominent publishers. Last, we present some results derived from this dataset that illustrate how the MBD can uncover correlates of bias in news coverage.