EMOTIONAL PATTERNS IN HEADLINES OF ENGLISH-LANGUAGE EUROPEAN MEDIA COVERAGE OF THE RUSSIA-UKRAINE WAR

Abstract

This study conducts automated emotional analysis of headlines from English-language European media outlets (DW, France 24, Politico Europe, Euronews, Der Spiegel International, El País English, RFI English, and Balkan Insight) reporting on the Russia-Ukraine war between February 2022 and October 2025. The significance of this research arises from the increasing focus on media discourse analysis and the challenges posed by contemporary information environments, including high information volumes and the fast pace of life, which hinder detailed source examination and diminish readers’ attention. Media narratives consequently influence societal perceptions and may reduce critical engagement with events. Emotionally charged vocabulary is frequently employed to capture attention and elicit empathy, and can also function as a tool for disinformation and emotional manipulation. The methodology involves constructing a text corpus and applying automated emotional analysis using the transformer-based emotion classification model, Emotion English DistilRoBERTa-base. Findings indicate that neutrality is the most prevalent category (46.43%), followed by anger (17.86%) and fear (17.86%). Disgust is less common (10.71%), while sadness is the least frequent among the identified emotions (7.14%). Joy and surprise were not detected in the analysed material. These results enhance understanding of the pragmatic mechanisms underlying media discourse, demonstrate the effectiveness of transformer-based models for natural language processing, and suggest prospects for future research, particularly in large-scale media narrative analysis.