Some difficulties in author attribution

E. S. Klyshinsky, N. A. Kochetkova, O. V. Karpik


Abstract: The paper investigates the issue of changing the relative position of texts written by different authors, while reducing the size of fragments of these works. To vectorize the text, we used statistics on the occurrence of syntactic connections, defined as parts of speech of words and the relation’s type. It is shown that decrease of the size of fragments leads to a blurring of the boundaries within which the works of one author are located.

Keywords: attribution of authorship, syntactic characteristics, corpus of texts, the Russian language, automatic text processing.

DOI: 10.20948/prepr-2024-84