Experiments in News Bias Detection with Pre-Trained Neural Transformers

TL;DR

This paper evaluates various pre-trained neural transformer models for detecting bias and classifying sub-types in news sentences, addressing the challenge of misinformation on the web.

Contribution

It systematically compares large pre-trained models for news bias detection and sub-type classification, providing both quantitative and qualitative insights.

Findings

Pre-trained transformers show promise in bias detection.

Model performance varies across bias sub-types.

Qualitative analysis highlights strengths and limitations.

Abstract

The World Wide Web provides unrivalled access to information globally, including factual news reporting and commentary. However, state actors and commercial players increasingly spread biased (distorted) or fake (non-factual) information to promote their agendas. We compare several large, pre-trained language models on the task of sentence-level news bias detection and sub-type classification, providing quantitative and qualitative results.