Визначення авторства україномовного тексту на основі методів машинного навчання

Abstract

The research focuses on the task of automatic authorship attribution for Ukrainian-language texts using machine learning methods. The proposed approach involves formalizing syntactic features obtained through parsing of sentences. A set of 11 structural characteristics has been developed to quantitatively represent an author’s style at the syntactic level. A multilayer perceptron is used to solve the attribution task, trained on the extracted features. Experimental results based on a corpus of texts by three Ukrainian authors confirm the effectiveness of the proposed method and demonstrate high classification accuracy with proper data preprocessing.