Development of the algorithm for graphematic analysis and isolating of semantically significant constructions in poorly structured text

Abstract

The article presents the results of analyzing the process of detecting patterns in semi-structured and unstructured texts. Text preprocessing is aimed at removing terms that do not carry useful information, bringing the words of the text to a single representation and forming a set of indices of a text information resource - statistical, linguistic, structural. An algorithm for graphematic analysis has been developed as the first stage of automatic processing of texts in natural language. The algorithm makes it possible to single out semantically significant natural language constructions in a semi-structured text using graphematic descriptors, to detect and replace abbreviations and abbreviations.

Hybrid client‐server implementation and microservice architecture of automatic documentation analysis software

“…To create a hierarchical data storage model in the developed software, it is proposed to use the object model [13].…”