Media characterization and landscape analysis

Research on characterizing and understanding the diverse landscape of news media through systematic analysis of source behavior, writing styles, editorial approaches, and engagement patterns. Encompasses quantitative measurement of linguistic and engagement features, comparative analysis of outlets across credibility and ideological spectrums, and mapping of the media ecosystem.

Key papers

• Horne et al. (2018) — Sampling the News Producers (NELA2017): Introduces systematic characterization of 92 news sources using 130 linguistic and engagement features, revealing systematic differences in writing style, sentiment, clickbait prevalence, and engagement patterns across mainstream, hyper-partisan, satire, and misinformation sources.
• Potthast et al. (2017) — A Stylometric Inquiry: Stylometric analysis distinguishing hyperpartisan from mainstream news using 1,627 manually fact-checked articles; demonstrates content-based style features (lexical, syntactic, structural) can characterize editorial bias and partisan lean.

Related topics

• News credibility assessment uses characterization features to evaluate source trustworthiness
• Fake news detection leverages source characterization as a signal for identifying misinformation
• Political polarization manifests in media landscape fragmentation that characterization studies reveal
• Bias detection focuses on the ideological and editorial dimensions of source characterization

Notes

Media characterization enables comparative analysis without requiring per-article ground truth labels — important for large-scale research where comprehensive fact-checking is infeasible. The field has evolved from binary outlets (reliable/unreliable) to multi-dimensional assessments capturing factuality, bias, transparency, and engagement.