Supplementary Material for 'From Articles to Publishers: Aggregating Language Model Predictions for News Source Reliability Inference'
This record contains the Supplementary Material for the manuscript "From Articles to Publishers: Aggregating Language Model Predictions for News Source Reliability Inference", submitted to IEEE Transactions on Computational Social Systems. Contents:- Appendix A: Detailed per-fold confusion matrices for BERT, RoBERTa, LLaMA 3, and Mistral- Appendix B: Hyperparameters, training curves, and aggregation strategies - Appendix C: Full list of 669 publishers with NewsGuard reliability scores and political orientation coding The supplementary PDF is referenced in the main manuscript and is provided to comply with the IEEE 10-page limit. It does not contain copyrighted news article texts.
Authors
- Fabio Pinelli (ORCID: https://orcid.org/0000-0003-1058-6917)
- John Bianchi (ORCID: https://orcid.org/0009-0006-2582-1480)
- Marinella Petrocchi (ORCID: https://orcid.org/0000-0003-0591-877X)
- Manuel Pratelli (ORCID: https://orcid.org/0000-0002-9978-791X)
Institutions
- IMT School for Advanced Studies Lucca (IT)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-06-18
- DOI
- https://doi.org/10.5281/zenodo.20744543
- Primary Topic
- Topic Modeling
- Type
- article
- Field-Weighted Citation Impact
- 0.00