Unveiling ideological extremes in parliamentary debates using transformer-based language models

Bibliographic Data

ID8134686
AuthorsBarbara Kos (0000-0001-5814-5454), Lana Tuković, Ema Vlainić, Marina Bagić Babac (0000-0003-4979-2216, corresponding author)
Year2026
Pages1-21
Publication date2026-01-19
Peer ReviewedYes
Open AccessNo
TypeARTICLE
VenueJournal of Information Technology & Politics (JOURNAL)
Journal identifiersISSN: 1933-169X • E-ISSN: 1933-1681
PublisherInforma UK Limited (PUBLISHER • GB)
DOI10.1080/19331681.2026.2616668
OpenAlexW7124952353
LanguageEN
References cited38

This study explores how natural language processing can uncover ideological orientations in European parliamentary debates. Using speeches from 28 countries, transformer-based models (BERT, RoBERTa, and DistilBERT) were fine-tuned for binary classification and evaluated by accuracy and F1-score. RoBERTa and DistilBERT slightly outperformed BERT, and DistilBERT offered a strong balance between efficiency and accuracy. Principal Component Analysis and attention score analysis revealed cross-national linguistic and ideological patterns. The findings advance understanding of political discourse through computational linguistics, highlighting the potential of transformer models for analyzing ideological diversity in large-scale parliamentary data.

Critical discourse analysis · Field (mathematics · Ideology · Key (lock · Politics · Computational and Text Analysis Methods · Populism, Right-Wing Movements · Sentiment Analysis and Opinion Mining

Tools

Open DOI