Ideology Prediction of German Political Texts

Hugging Face Daily Papers Papers

Summary

This paper presents a transformer-based model that projects political orientation of German texts onto a continuous left-to-right spectrum, achieving high accuracy across multiple corpora including Bundestag plenary notes, Wahl-O-Mat, newspapers, and tweets.

Elections represent a crucial milestone in a nation's ongoing development. To better understand the political rhetoric from various movements, ranging from left to right, we propose a transformer-based model capable of projecting the political orientation of a text on a continuous left-to-right spectrum, represented by a normalized scalar d between -1 and 1. This approach enables analysts to focus on specific segments of the political landscape, such as conservatives, while excluding liberal and far-right movements. Such a task can only be achieved with multiclass classifiers, provided that the desired orientation is incorporated within one of their predefined classes. To determine the most suitable foundation model among 13 candidate transformers for this task, we constructed four distinct corpora. One corpus comprised annotated plenary notes from the German Bundestag, while another was based on an official online decision-making tool, Wahl-O-Mat. The third corpus consisted of articles from 33 newspapers, each identified by its political orientation, and the fourth included 535,200 tweets from 597 members of the 20th and 21st German Bundestag. To mitigate overfitting, we used two distinct corpora for training and two for testing, respectively. For in-domain performance, DeBERTa-large achieved the highest F1 score F1=0.844 as well as for the X (Twitter) out-of-domain test ACC=0.864. Regarding the newspaper out-of-domain test, Gemma2-2B excelled (MAE = 0.172). This study demonstrates that transformer models can recognize political framing in German news at the level of public opinion polls. Our findings suggest that both the model architecture and the availability of domain-specific training data can be as influential as model size for estimating political bias. We discuss methodological limitations and outline directions for improving the robustness of bias measurement.
Original Article
View Cached Full Text

Cached at: 05/15/26, 04:24 AM

Paper page - Ideology Prediction of German Political Texts

Source: https://huggingface.co/papers/2605.14352

Abstract

A transformer-based model projects political orientation on a continuous spectrum using multiple corpora, achieving high accuracy in detecting political bias across different text sources.

Elections represent a crucial milestone in a nation’s ongoing development. To better understand the political rhetoric from various movements, ranging from left to right, we propose atransformer-based modelcapable of projecting thepolitical orientationof a text on a continuousleft-to-right spectrum, represented by a normalized scalar d between -1 and 1. This approach enables analysts to focus on specific segments of the political landscape, such as conservatives, while excluding liberal and far-right movements. Such a task can only be achieved withmulticlass classifiers, provided that the desired orientation is incorporated within one of their predefined classes. To determine the most suitablefoundation modelamong 13 candidate transformers for this task, we constructed four distinct corpora. One corpus comprised annotated plenary notes from the German Bundestag, while another was based on an official online decision-making tool, Wahl-O-Mat. The third corpus consisted of articles from 33 newspapers, each identified by itspolitical orientation, and the fourth included 535,200 tweets from 597 members of the 20th and 21st German Bundestag. To mitigate overfitting, we used two distinct corpora for training and two for testing, respectively. Forin-domain performance,DeBERTa-largeachieved the highestF1 scoreF1=0.844 as well as for the X (Twitter)out-of-domain testACC=0.864. Regarding the newspaperout-of-domain test,Gemma2-2Bexcelled (MAE= 0.172). This study demonstrates that transformer models can recognize political framing in German news at the level of public opinion polls. Our findings suggest that both the model architecture and the availability of domain-specific training data can be as influential as model size for estimating political bias. We discuss methodological limitations and outline directions for improving the robustness of bias measurement.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.14352

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.14352 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.14352 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.14352 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Ideology Prediction of German Political Texts

arXiv cs.CL

The paper proposes a transformer-based model to predict political ideology of German political texts on a continuous left-to-right spectrum. The study compares 13 models and finds DeBERTa-large and Gemma2-2B perform best on different tasks.

TextLDM: Language Modeling with Continuous Latent Diffusion

Hugging Face Daily Papers

This paper introduces TextLDM, a method that adapts visual latent diffusion transformers for language modeling by mapping discrete tokens to continuous latents. It demonstrates that this approach, enhanced by representation alignment, matches GPT-2 performance and unifies visual and text generation architectures.

Polarization by Default: Auditing Recommendation Bias in LLM-Based Content Curation

arXiv cs.CL

This paper presents a large-scale audit of recommendation biases in LLM-based content curation across OpenAI, Anthropic, and Google using 540,000 simulated selections from Twitter/X, Bluesky, and Reddit data. The study finds that LLMs systematically amplify polarization, exhibit distinct toxicity handling trade-offs, and show significant political leaning bias favoring left-leaning authors despite right-leaning plurality in datasets.