Development and Training of a Language Model for Automated Classification of English-language Scientific Articles by SRSTI Codes

In the modern scientific community there is a growing need for effective classification of scientific articles to improve the search and organization of knowledge due to the rapid growth of information. The purpose of the graduate qualification work is to develop a language model based on artificial intelligence to automate the process of classification of English-language scientific articles by GRSTI codes. The relevance of the project is due to the need to improve the accuracy and speed of scientific data processing, as currently the classification process is carried out manually. The aim of the study is to create a model capable of automatically determining the appropriate GRSTI codes for articles, which will improve the process of reviewing and indexing publications. The research methods include using the collected corpus of articles, its preprocessing, developing a special architecture to solve the hierarchical classification problem, training the model using machine learning algorithms and evaluating its effectiveness. The result of this research is the developed console application for automatic classification of large volumes of scientific articles. The paper consists of 35 pages, 32 sources, 8 figures and 5 tables.