Authorship Attribution with Convolutional Neural Networks and POS-Eliding

Julian Hitschler, Esther van den Berg, Ines Rehbein
2017 Proceedings of the Workshop on Stylistic Variation   unpublished
We use a convolutional neural network to perform authorship identification on a very homogeneous dataset of scientific publications. In order to investigate the effect of domain biases, we obscure words below a certain frequency threshold, retaining only their POS-tags. This procedure improves test performance due to better generalization on unseen data. Using our method, we are able to predict the authors of scientific publications in the same discipline at levels well above chance.
doi:10.18653/v1/w17-4907 fatcat:y5tvbnwzuferjh3ogbv4673mcq