Rethinking the Authorship Verification Experimental Setups

Florin Brad, Andrei Manolache, Elena Burceanu, Antonio Barbalau, Radu Tudor Ionescu, Marius Popescu


Abstract
One of the main drivers of the recent advances in authorship verification is the PAN large-scale authorship dataset. Despite generating significant progress in the field, inconsistent performance differences between the closed and open test sets have been reported. To this end, we improve the experimental setup by proposing five new public splits over the PAN dataset, specifically designed to isolate and identify biases related to the text topic and to the author’s writing style. We evaluate several BERT-like baselines on these splits, showing that such models are competitive with authorship verification state-of-the-art methods. Furthermore, using explainable AI, we find that these baselines are biased towards named entities. We show that models trained without the named entities obtain better results and generalize better when tested on DarkReddit, our new dataset for authorship verification.
Anthology ID:
2022.emnlp-main.380
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5634–5643
Language:
URL:
https://aclanthology.org/2022.emnlp-main.380
DOI:
Bibkey:
Cite (ACL):
Florin Brad, Andrei Manolache, Elena Burceanu, Antonio Barbalau, Radu Tudor Ionescu, and Marius Popescu. 2022. Rethinking the Authorship Verification Experimental Setups. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 5634–5643, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Rethinking the Authorship Verification Experimental Setups (Brad et al., EMNLP 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingestion-script-update/2022.emnlp-main.380.pdf