Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Connecting the Dots in News Analysis: Bridging the Cross-Disciplinary Disparities in Media Bias and Framing

Gisela Vallejo, Timothy Baldwin, Lea Frermann

TL;DR

This paper addresses the gap between social science theories of framing and media bias and contemporary NLP approaches. It surveys NLP and social science literature to identify three core disconnects: local versus global framing signals, dynamic changes in framing over time and contexts, and the value of cross-document comparative analysis. The authors propose concrete directions, including multi-level and event-centric annotations, cross-document grounding with external information, and treating bias as a comparative task, along with emphasis on transparency and open data. By aligning NLP methods with social science methodologies, the work aims to produce scalable yet theory-informed analyses of media bias with practical implications for researchers, journalists, and the public.

Abstract

The manifestation and effect of bias in news reporting have been central topics in the social sciences for decades, and have received increasing attention in the NLP community recently. While NLP can help to scale up analyses or contribute automatic procedures to investigate the impact of biased news in society, we argue that methodologies that are currently dominant fall short of addressing the complex questions and effects addressed in theoretical media studies. In this survey paper, we review social science approaches and draw a comparison with typical task formulations, methods, and evaluation metrics used in the analysis of media bias in NLP. We discuss open questions and suggest possible directions to close identified gaps between theory and predictive models, and their evaluation. These include model transparency, considering document-external information, and cross-document reasoning rather than single-label assignment.

Connecting the Dots in News Analysis: Bridging the Cross-Disciplinary Disparities in Media Bias and Framing

TL;DR

This paper addresses the gap between social science theories of framing and media bias and contemporary NLP approaches. It surveys NLP and social science literature to identify three core disconnects: local versus global framing signals, dynamic changes in framing over time and contexts, and the value of cross-document comparative analysis. The authors propose concrete directions, including multi-level and event-centric annotations, cross-document grounding with external information, and treating bias as a comparative task, along with emphasis on transparency and open data. By aligning NLP methods with social science methodologies, the work aims to produce scalable yet theory-informed analyses of media bias with practical implications for researchers, journalists, and the public.

Abstract

The manifestation and effect of bias in news reporting have been central topics in the social sciences for decades, and have received increasing attention in the NLP community recently. While NLP can help to scale up analyses or contribute automatic procedures to investigate the impact of biased news in society, we argue that methodologies that are currently dominant fall short of addressing the complex questions and effects addressed in theoretical media studies. In this survey paper, we review social science approaches and draw a comparison with typical task formulations, methods, and evaluation metrics used in the analysis of media bias in NLP. We discuss open questions and suggest possible directions to close identified gaps between theory and predictive models, and their evaluation. These include model transparency, considering document-external information, and cross-document reasoning rather than single-label assignment.
Paper Structure (25 sections, 2 figures, 2 tables)

This paper contains 25 sections, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Contribution and Approach
  3. Background: Framing and Media Bias
  4. Three Disconnects
  5. Framing/bias is local and global
  6. Framing/bias is dynamic
  7. Framing/bias as a comparative task
  8. A Critical Review of Current Practices in NLP and Social Science
  9. Datasets
  10. Media bias detection
  11. Framing
  12. Methodologies
  13. Opportunities for Future Work.
  14. Evaluation
  15. Validating annotations
  16. ...and 10 more sections

Figures (2)

  • Figure 1: Two articles about the same event written from different political ideologies (Source: allsides.com).
  • Figure 2: Illustration of the three disconnects: framing is both local and global (blue), dynamic (green) and best identified through comparative analysis (yellow).