From Texts to Knowledge Graph in the Semantic Library LibMeta

Abstract

The problem of constructing a knowledge graph for previously unprocessed scientific texts is studied. The task is to analyze the use of various machine processing methods for extracting text semantics using the example of an interdisciplinary subject area. The goal of the research is to create a semantic model of the subject area of an interdisciplinary scientific journal and use a knowledge graph to navigate through data. The semantic model is based on the ontology of the LibMeta library. Texts and metadata receive connections due to the processing procedure and become part of the ontology of the semantic library. The interdisciplinary subject area is integrated into the library content as a related area of the ‘‘mathematics’’ ontology of the LibMeta library. A comparison of texts is carried out according to different characteristics, thematic proximity, commonality of research methods and approaches, problem setting and their symbolic representation. The results of preliminary text processing and the results in the form of a knowledge graph integrated into the library content are presented. The formulation of the problem of constructing an ontology and a knowledge graph for an interdisciplinary subject area adjacent to mathematics is considered by the authors as part of the general problem of managing mathematical knowledge.