In this project we investigate different approaches for the task of "Authorship Verification", highlighting differences in terms of performances. This task, originally presented at PAN 2020 conference, consists in the evaluation of different fictional texts in order to see if they have been written by the same author using Natural Language Processing techniques, which involve also neural approaches. Please note that the repository doesn't contains the datasets, that can be requested here.
You can see the full report here.
The project has been developed by Gabriele Pisciotta, Miriana Somenzi and Fabio Murgese, in the context of the Human Language Technologies course at the University of Pisa, held by professor Attardi.