An architecture of a system for the semantic analysis of the scientific and technical documents

A. V. Kan^ab, Alexander A. Khoroshilov^ab, I. A. Chechulin^a, S. A. Stupnikov^ca, Yu. V. Nikitin^d, Alexey A. Khoroshilov<sup>a

a</sup> All-Russian Institute for Scientific and Technical Information of the Russian Academy of Sciences, 20 Usievicha Str., Moscow 125190, Russian Federation
^b Moscow State Aviation Institute (National Research University), 4 Volokolamskoe Shosse, Moscow 125933, Russian Federation
^c Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
^d National Research Center "Zhukovsky Institute," 1 Zhukovsky Str., Zhukovsky 140180, Russian Federation

Abstract: The paper considers the functional architecture of the system of semantic analysis of scientific and technical documents. The system belongs to the category of intelligent platforms for the aggregation and analysis of scientific and technical documents. The main technological solution of the system is to create a set of declarative tools adaptable to various domains of science and technology, implementing the concept of phraseological conceptual analysis of texts. This assumes combining the results of graphemic, morphological, conceptual, and semantic and syntactic analysis of texts into a formalized metadata model. The subsystems and modules constituting the architecture are considered, their functions and interaction are described. One of the most important components of the system is a set of morphological, semantic, syntactic, and conceptual dictionaries that are automatically generated and customized to domains of knowledge, ensuring the relevance of a set of linguistic resources. A classification of systems for semantic analysis of texts and the problems they solve are considered. The system being developed is positioned in the classification and its distinctive features are distinguished.

Keywords: phraseological conceptual analysis of texts, semantic analysis of texts, scientific and technological text processing.

Received: 29.07.2025
Accepted: 15.10.2025

DOI: 10.14357/08696527250404