Reproducibility and Robustness of Authorship Identification Approaches

Warning

This publication doesn't include Institute of Computer Science. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

KARÁSEK Adam NEVĚŘILOVÁ Zuzana

Year of publication 2023
Type Article in Proceedings
Conference Proceedings of the Seventeenth Workshop on Recent Advances in Slavonic Natural Languages Processing, RASLAN 2023
MU Faculty or unit

Faculty of Informatics

Citation
Web
Keywords authorship identification; evaluation; reproducibility
Description Authorship identification, framed as a classification task, assigns a digital text to a known author. State-of-the-art algorithms for this task often lack evaluation across diverse datasets. This paper reimplements and evaluates three approaches on three different datasets, exploring the robustness of algorithms on various text types (e.g., emails, articles, instant messages). Not all the published methods are fully reproducible. However, reasonable parameters were selected if they were not part of the original paper. The evaluation of the ensemble model shows it is somewhat robust on different texts and different numbers of potential authors.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.

More info