Scientific publications

This paper proposes an approach to the quantitative analysis of the social structure of Russian literary communities by constructing a graph of personal connections among writers. The relevance of the study arises from the lack of structured datasets in this subject domain and the growing significance of digital humanities methodologies for investigating the sociodynamics of literature. The empirical basis for constructing a mathematical community model was the list of authors from the official “Unified State Exam Codifier in Literature” (2026). Thissource ensures representativeness while avoiding artificial connectedness among authors. Social connections data, including personal communication, correspondence, and membership in professional unions, were collected for 102 Russian writers active between 1757 and 2003. The data collection process employed neural network models (DeepSeek, YandexGPT) followed by expert verification. The results showed that the proposed method for constructing the social interactions graph yields results consistent with expert data and can be used for sociodynamic analysis of literary communities. Social network analysis methods are well suited for studying writer communities and offer an alternative approach to assessing a writer’s status through the structure of personal connections.