Year: 2025
Type: conference-paper
Abstract: We used a novel methodology for analyzing peace representations in global media by integrating a Retrieval-Augmented Generation (RAG) model with refined concepts of Positive and Negative Intergroup Reciprocity (PIR/NIR). Traditional analyses often overlook nuanced intergroup dynamics. By refining PIR and NIR definitions, we provide a precise framework to assess media portrayals of intergroup relations influencing peace. Using the News On the Web dataset of 700,000 articles from 18 countries, we employed machine learning to embed articles and identify PIR and NIR instances. This allowed us to classify countries based on their media’s alignment with these concepts. Our findings reveal significant variations, with some media emphasizing intergroup tolerance and others highlighting aggression. Comparing our PIR/NIR analysis with traditional peace indices, we demonstrate that our approach offers deeper insights into factors influencing national peace. This has significant implications for peace building, high lighting the critical role of media in shaping peace.
Authors: Kevin Lian, Larry S. Liebovitch, Melissa Wild, Harry West, Peter T. Coleman, Francine Chen, Everlyne Kimani, Kate Sieck
Institutions: Columbia University, Toyota Research Institute
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
Funder: Toyota Research Institute
OpenAlex ID: W4409156149