Integrating the subject area subspace into the “mathematics” semantic space

The paper studies the problem of developing a semantic library by adding a new applied scientific area. The authors use the example of a journal on applied issues of composite materials in order to build an addition to the main content of the library, to expand the description of the original subject area, and to detail UDC and MSC articles corresponding to the local subject area. At the same time, they solve the tasks of adding terms to the thesaurus, building a reference corpus of the applied subject area of mathematics, and creating a custom interface. Formulas and equations of the local subject area are semantically linked to the main content of the library. The main advantage of using semantic libraries for this kind of tasks is to enrich the existing knowledge base of the library and identify relationships in data. To study these problems, it is necessary to interact with subject matter experts and use modern tools and methods for natural language processing, machine learning approaches to knowledge representation. The above tasks are implemented by means of the LibMeta semantic library, which allows considering such concepts as the thematic area/subdomain and their hierarchical relationships, in particular, the automatic inclusion of the enclosing area concepts in the subdomain knowledge system; the possibility of personalizing the built structures and using LibMeta as a construction tool, in particular, building the ontology itself. Data integration within the library allows expanding the description of subject areas related to the applications of mathematics in interdisciplinary research and technology. As an example of one of the applied sections of mathematical physics problems, the procedure for including specific dictionaries, thesauri, as well as an array of publications of a specialized journal in a semantic library is shown. The proposed approach will allow using the content of the “mathematics” semantic library for scientific research and minimizing the process of searching for information in the local subject area, without losing more general results contained outside this area.