EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Unified Metadata-Aware Transformer Framework for Misleading News Classification

Mahmudul Haque, A. S. M. Hossain Bari, Marina L. Gavrilova

IEEE Access · 2026 · 14 · 116304-116320

DOI: 10.1109/access.2026.3717645 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Mahmudul Haque und 2 weiteren Autor:innen. Die Arbeit nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant, Alberta Innovates, New Frontiers in Research Fund.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Misleading information appearing on various digital platforms poses serious risks to public trust. Recently developed approaches to misleading news detection rely primarily on binary classification; however, real-world fact-checking scenarios demand multiclass formulations that capture varying degrees of truthfulness. To address this gap, our research introduces MetaRoBERTa, a unified transformer-based representation learning framework, as a robust solution for multiclass misleading news classification. MetaRoBERTa integrates claims, contextual information, justification text, and speaker-related credibility metadata into a single textual representation, enabling end-to-end learning without explicit linguistic or behavioral feature engineering. The proposed approach achieves an accuracy of 80.44% on the semantically rich LIAR2 dataset, when justification text is available, surpassing state-of-the-art results. Ablation studies further show that the complete regularized training strategy yields an improvement over a cross-entropy baseline that is nominally significant at the 0.05 level before correction for multiple comparisons, with individual components contributing small, consistently positive gains, and per-class analysis reveals dataset-dependent behavior that provides practical guidance for model selection and deployment.

Förderung

  • Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant
  • Alberta Innovates
  • New Frontiers in Research Fund
  • University of Calgary Transdisciplinary Scholarship Connector Grant
  • University of Calgary Research Excellence Chair in Trustworthy and Explainable Artificial Intelligence

Zitieren

APA

Haque, M., Bari, A. S. M. H., & Gavrilova, M. L. (2026). Unified Metadata-Aware Transformer Framework for Misleading News Classification. IEEE Access, 14, 116304-116320. https://doi.org/10.1109/access.2026.3717645

Vancouver

Haque M, Bari ASMH, Gavrilova ML. Unified Metadata-Aware Transformer Framework for Misleading News Classification. IEEE Access. 2026;14:116304-116320. doi:10.1109/access.2026.3717645

BibTeX
@article{Haque2026,
  author = {Haque, Mahmudul and Bari, A. S. M. Hossain and Gavrilova, Marina L.},
  title = {Unified Metadata-Aware Transformer Framework for Misleading News Classification},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {116304--116320},
  doi = {10.1109/access.2026.3717645}
}
Kostenlos starten →