1. Home
2. Archive
3. 2025
4. №1(61), 2025
5. Genre classification of literary texts through neural network methods (based on the russian-language fanfiction electronic database)

Genre classification of literary texts through neural network methods (based on the russian-language fanfiction electronic database)

Maksimenko Polina Igorevna

Intern Researcher, Language Convergence Laboratory, National Research University Higher School of Economics, Russia, Saint-Petersburg, p.maksimenko@hse.ru

Abstract

The article focuses on the fine-tuning and application of a large language model based on BERT for genre classification of literary texts. The training data for the neural network algorithm were created based on fanfiction works from a Russian-language fanfiction electronic database containing more than 160,000 texts. The training dataset for the neural network algorithm contains fanfiction texts, each of which is marked with one of eight genre labels or more. The article presents the results of testing and evaluation of the effectiveness of three genre classifier ver-sions (multi-label, multi-class, and binary) on both the original dataset and a test sample of literary fiction. The research also includes the comparison of classifica-tion quality values for texts of different genres.

Keywords

fanfiction; genre classification; mass literature; BERT.

For citing: Maksimenko P.I. (2025) Genre classification of literary texts through neural network methods (based on the russian-language fanfiction electronic database). Human being: Image and essence. Humanitarian aspects. Moscow. INION RAN.Vol. 1 (61). pp. 184-200. DOI: 10.31249/chel/2025.01.13

Order this article