An automated system for key terms analysis

The paper describes the method proposed by the authors for forming an array of key terms that form the basis of subject ontology of the Common Digital Space of Scientific Knowledge and encyclopedias list of articles (slots) in a certain scientific direction. The method is based on a frequency analysis of the occurrence of key terms in the articles published in leading scientific journals on this topic. The technique involves program processing of metadata of articles reflected in various databases, constructing rating lists of the frequency of occurrence of individual key terms and selecting the cores of such lists, which, in turn, can be considered as the basis for filling a subject ontology and forming encyclopedic slots. To implement the methodology, the authors developed the corresponding database structure, software tools for filling it, processing and analyzing data. The paper presents the database description and the results of the practical implementation of the methodology based on processing several thousand articles from leading Russian journals in mathematics, informatics and physics (terms in Russian and English were identified and analyzed). An assessment of the correspondence of key terms distribution frequencies and their constituent individual words to the Bradford law has shown significant discrepancies with this law in the case of key terms, but there was certain convergence when considering individual words and their permutation within key terms.