Emotion-Driven Summarization: Exploring How Emotional Rhetoric in Political Communication Distorts Automated and Human Summaries

Building on Li & Zhai's (2025) findings about the strategic use of emotional and propagandistic rhetoric in political social media, this idea explores a novel angle: when political content is highly emotional, do summaries (by humans or machines) tend to amplify, dampen, or selectively transmit those emotions? This would involve content analysis and sentiment tracking across the original messages and their summaries, comparing both human-edited and AI-generated outputs. The research could test competing hypotheses: that summarization acts as a filter (reducing emotional intensity), as an amplifier (heightening affective framing), or as a selective transmitter (emphasizing certain emotions over others). Such work would connect cognitive psychology (emotion contagion), media studies, and computational linguistics, and could inform guidelines for more responsible summarization in politically sensitive contexts. The implications for media ethics, audience polarization, and AI design are substantial.

References:

1. Strategic Emotional Rhetoric in Political Social Media: A Comprehensive Analysis of Donald Trump’s Tweeting Patterns and Propagandistic Techniques in the Digital Age. Yiping Li, Yuejing Zhai (2025). Studies in Linguistics and Literature.