Multispecies Ensemble Architecture For Texts Authorship Classification

This paper deals with the problem of text authorship identification. This issue is now one of the most demanded tasks in the field of natural language processing. A review of popular authorship analysis techniques is conducted. The implementation of a hybrid classifier built on the ensemble approach is suggested in the paper. By balancing a variety of methodologies and choosing the best outcomes from them, this is anticipated to produce more accurate authorship identification quality measures. This classifier offers a consistently high level of author identification. The development of an ensemble approach with a quantum-inspired component is offered as a classifier architecture. A number of experiments have verified the method's benefits. The results of computer experiments are presented, which confirmed the hypotheses about the effectiveness of the ensemble approach to the task of text classification. The proposed classifier can be effectively applied to identify borrowing cases. For example, in an educational environment, to better detect student plagiarism. At the moment, the applied implementation of the classifier is being tested in the form of a support module for the University Academic Integrity System.