CONTOUR: A Framework for Investigating LLM-Generated and Human-Written Knowledge on Controversial Topics

ACL ARR 2026 January Submission9485 Authors

Keywords: NLP tools for social analysis, evaluation and metric, Large Language Model

Abstract: As Large Language Models (LLMs) increasingly serve as primary information-seeking tools, their role in shaping public perception of complex social issues becomes a critical concern for civic discourse. This study investigates the representation of controversial topics in LLMs. We propose a comparative framework that benchmarks LLM-generated content against two human-curated knowledge systems: Wikipedia, representing community-driven consensus, and Encyclopedia Britannica, representing elite expert viewpoints. Through a multi-dimensional linguistic analysis across 153 politically sensitive topics, we quantify the "intensity of controversy" using a novel taxonomy.

Paper Type: Short

Research Area: Computational Social Science, Cultural Analytics, and NLP for Social Good

Research Area Keywords: NLP tools for social analysis, evaluation and metric

Contribution Types: NLP engineering experiment, Reproduction study, Publicly available software and/or pre-trained models, Data resources

Languages Studied: English

Submission Number: 9485