Classifying Peace in Global Media Using RAG and Intergroup Reciprocity

TL;DR

The paper addresses how peace is reflected in global media and challenges traditional sentiment-based analyses by introducing a Retrieval-Augmented Generation (RAG) framework to quantify Positive Intergroup Reciprocity ($PIR$) and Negative Intergroup Reciprocity ($NIR$) at the national level. It embeds article content from the NOW dataset into $1536$-dimensional vectors, stores them in ChromaDB, and uses a RAG pipeline that blends social science theory with media content to identify PIR and NIR patterns. The study refines PIR/NIR definitions, demonstrates a cross-national analysis across 18 countries, and provides concrete examples (eg, Singapore as PIR, Kenya as NIR) along with a comparative PIR alignment metric. This approach offers a real-time, qualitative measure of peace dynamics in media that complements traditional indices like PPI and HDI and has practical implications for policy and peacebuilding by highlighting how media portrayals can influence social cohesion.

Abstract

This paper presents a novel approach to identifying insights of peace in global media using a Retrieval Augmented Generation (RAG) model and concepts of Positive and Negative Intergroup Reciprocity (PIR/NIR). By refining the definitions of PIR and NIR, we offer a more accurate and meaningful analysis of intergroup relations as represented in media articles. Our methodology provides insights into the dynamics that contribute to or detract from peace at a national level.

Figures (1)

• Figure 1: Workflow of the RAG implementation process.