Using Machine Learning Methods for the Analysis of Mass Historical Sources as Exemplified by Letters from the Great Patriotic War

The research is devoted to the analysis of the corps of letters from the time of the Great Patriotic War. The relevance of the research topic is determined by the need to introduce into the scientific circulation digitized front-line epistolary sources with valuable information potential. The study of front-line letters helps to understand many aspects of front-line experience, the objective and holistic reconstruction of which is possible, including through the expansion of the source field. The significance of the topic under consideration is also determined by the need to improve the methodology for analyzing mass historical sources using modern technologies. The study solves the following tasks: 1. To analyze the literature and research experience on the analysis of mass historical sources; 2. To study the linguistic techniques and methods of natural language processing used in the digital humanities; 3. Introduce into scientific circulation a digital collection of letters from the Great Patriotic War using the international standard TEI; 4. Develop a methodology based on which to analyze private front-line correspondence using machine learning methods.