Today most media content is consumed based on algorithmic recommendations. Evidence suggests that this can lead to politically biased media consumption patterns. Automated extraction of political agendas from texts can reveal and analyze political biases in online media -- and thus help fostering politically unbiased media consumption. Here we employ modern political text analysis methods demonstrating the potential of automated fine-grained political bias analysis in online media. We conduct an analysis of political content in German language online media during the period 2019--2022, encompassing several million articles and tweets covering events with profound societal impact globally and nationally, the COVID-19 pandemic and the beginning of the war in Ukraine. Our analysis identifies thematic similarity between national (German and Swiss) reporting, particularly for categories driven by international events. We also find divergences emerging in domestically influenced categories, reflecting differences in national policies and institutional structures. A comparison of newspaper and Twitter discourse reveals that both media converge around a shared core during the pandemic, yet differ in intensity and temporal dynamics. Newspapers exhibit more stable political content, while Twitter reacts through short-lived event-driven spikes. These findings indicate that international crises act as a powerful synchronizing force on political content in classical media, temporarily overriding both national and media-form differences. Our automated political analysis empowers citizens by rendering political agendas in online media transparent. This transparency also enables media outlets to bridge the gap between algorithm-driven echo chambers and a more informed, balanced public discourse.
Figures & tables
Year
Germany
Switzerland
2019
489,162
2,824
2020
655,012
31,705
2021
784,272
30,675
2022
1,195,600
32,729
Total
3,124,046
97,933
Table 1: Number of crawled news articles by year and country.
Figure 1: Distribution of articles across newspapers in Germany (DE) and Switzerland (CH). The dataset is dominated by German outlets, with only three Swiss newspapers included: Tages-Anzeiger, Schweizer Radio und Fernsehen, and Neue Zürcher Zeitung (NZZ).
Figure 2: Longitudinal trends in thematic domain shares across newspapers and Twitter across pre-pandemic, pandemic core, and post-pandemic phases. Newspapers exhibit relatively stable distributions throughout, while Twitter shows considerably higher volatility, particularly during the pandemic core phase, where several domains spike sharply before subsiding.
Figure 3: Distribution of thematic domains in newspapers and Twitter across three pandemic phases (pre-pandemic, pandemic-core, post-pandemic), revealing a convergence in topic coverage between traditional and social media during COVID-19.
Figure 4: Distribution of individual thematic labels in newspapers and Twitter across pandemic phases. In contrast to the overarching domain level, individual labels exhibit pronounced differences between the two media, with Twitter showing more volatile and extreme shifts in topic salience.
Categories
Correlation
105 - Military: Negative
0.93
504 - Welfare State Expansion
0.89
303 - Governmental and Administrative Efficiency
0.85
505 - Welfare State Limitation
0.85
103 - Anti-Imperialism
0.85
106 - Peace
0.83
Table 2: Top and bottom thematic correlations of monthly ratios (DE vs. CH)
Figure 5: Monthly reporting ratios in German (DE) and Swiss (CH) news papers for the topics Federalism ( ρ=0.34 ) and Education Expansion ( ρ=0.09 ) show the lowest correlations in the dataset. This could reflect national differences such as more direct participatory democracy in Switzerland and differences in COVID-19 measures in schools.
Figure 6: Comparative analysis of Welfare State Expansion, Military, and Democracy reporting in Germany and Switzerland (2020–2022). Both countries show closely aligned trends in Military: Negative and Welfare State Expansion , reflecting shared responses to global events. In contrast, Democracy reveal pronounced national differences.
Figure 7: Monthly share of left-wing and right-wing topics in the public german discourse, classified based on the Manifesto Project coding scheme. Across both countries, left-wing and right-wing shares remain relatively stable and balanced throughout the observation period. A notable exception occurs in early 2022, where a pronounced spike in left-wing topics is observable in both countries, coinciding with the Russian invasion of Ukraine.
Social media platforms have become central to shaping political discourse, serving as arenas where narratives form and evolve, influencing public opinion. Identifying and analyzing these narratives, particularly when they compete across different political ideologies, is crucial for understanding the dynamics of modern political communication. This paper presents an unsupervised framework for identifying and characterizing competing narratives in political discourse on social media, focusing on German politicians' tweets. The framework employs a multi-stage pipeline that integrates natural language processing techniques such as topic modeling, event detection, and event linking. By forming data into coherent stories and uncovering the distinct perspectives of user communities, the system is able to detect the key competing narratives, highlighting the divergent framings and conflicts surrounding trending political topics. Two case studies on polarizing political issues demonstrate the efficacy of the methodology, showcasing its ability to uncover and analyze divergent viewpoints. The findings contribute to the broader understanding of how narratives propagate within the digital public sphere and offer insights for policymakers, social media platforms, and researchers interested in monitoring political discourse.
Sergej Wildemann, Erick Elejalde
L3S Research Center, Leibniz Universität Hannover, Hannover, Germany
Wales' political landscape has been marked by growing accusations of bias in Welsh media. This paper takes the first computational step toward testing those claims by examining Nation.Cymru, a prominent Welsh political news outlet. I use a two-stage natural language processing (NLP) pipeline: (1) a robustly optimized BERT approach (RoBERTa) bias detector for efficient bias discovery and (2) a large language model (LLM) for target-attributed sentiment classification of bias labels from (1). A primary analysis of 15,583 party mentions across 2022-2026 news articles finds that Reform UK attracts biased framing at twice the rate of Plaid Cymru and over three times as negative in mean sentiment (p<0.001). A secondary analysis across four parties across both news and opinion articles shows that Plaid Cymru is the outlier, receiving markedly more favourable framing than any other party. These findings provide evidence of measurable differential framing in a single Welsh political media outlet, supporting calls for a broader review of Welsh media coverage. Furthermore, the two-stage pipeline offers a low-cost, replicable framework for extending this analysis to other Welsh outlets, as well as media ecosystems outside of Wales.
Cai Parry-Jones
University of Cambridge, Cambridge, United Kingdom
This paper presents a pipeline for analyzing media bias and framing in online news. The pipeline groups articles into topics and events, adds named-entity and sentiment annotations, and compares news sources through people mentions, source-level tone, and event-level coverage patterns. We apply it to 8,358 Albanian news articles collected from GDELT and compare the resulting annotations with GDELT's automated annotations. The results show moderate agreement for sentiment and entity extraction, as well as additional person-entity pairs that can potentially support the bias analysis. We compare two annotation prompts and find that stricter sentiment-validation rules remove label-score inconsistencies but increase execution time and reduce annotation coverage. Based on these results, the simpler prompt is used for the rest of the analysis. We have provided sample analysis on source-level framing pro les, person-level tone differences across sources, and event-level gatekeeping and coverage indicators. These outputs show how the same news collection can be used to examine what sources cover, how they describe public figures, and where coverage is concentrated. The approach is particularly useful in settings where manually verified datasets or specialized language tools are limited.
Klesti Hoxha, Olti Qirici
Department of Informatics Faculty of Natural Sciences University of Tirana Tirana, Albania