Creation of a Biographical Dataset of Russian Prose Writers of the 20th Century Using ChatGPT-4o

Abstract

In the 21st century, a rapid advancement of digital technologies has given rise to new methodological approaches within the humanities, including literary studies and biographical research. The automation of data processing and information annotation is increasingly supported by Large Language Models (LLMs), such as ChatGPT. The emergence of these tools has prompted a shift from traditional methodologies based exclusively on manual annotation and expert verification toward hybrid strategies that combine automated information extraction with subsequent human curation. Such an approach enhances the efficiency and sustainability of data collection and processing while preserving a high level of factual accuracy. LLMs demonstrate significant potential for automatic information extraction and classification, contributing to more systematic and scalable approaches to data-driven research in the humanities. The aim of this study is to create a biographical dataset of 20th-century Russian prose writers using ChatGPT-4o. The dataset structure, comprising 30 categories describing various aspects of the authors’ personal lives and creative activities, was defined by experts at the initial stage of the study. The model received prompts designed to formulate and iteratively refine user-defined tasks. Several techniques for constructing effective prompts aimed at improving the quality and precision of model responses are described in the paper. Automatically generated data were manually verified, and all model errors corrected. The testing demonstrated that ChatGPT-4o can be effectively employed for the automated preparation and preliminary analysis of large volumes of textual data; it also revealed the model’s main strengths and limitations in extracting and structuring biographical information. The resulting dataset, represented as a relational table containing biographical data for 650 Russian writers, became part of the biographical module of the Corpus of the 20th Century Russian Short Stories. Its structure enables interdisciplinary research in the fields of literary sociology and modeling of the national literary process, as well as the study of correlations between authors’ biographical features and the artistic, thematic, and linguistic characteristics of their works.