Search academic texts

Registration Number

0221U100983, 0120U103416 , R & D reports

Title

Develop a personalized information system to support scientific research in the institutions of NAS of Ukraine. Stage 1. Develop a system of basic services of semantic analysis of natural language description of defined subject fields.

popup.stage_title

Head

Palagin Oleksandr V., Доктор технічних наук

Registration Date

11-01-2021

Organization

V.M. Glushkov Institute of Cybernetics of National Academy of Sciences of Ukraine

popup.description2

Object of the research: methods of construction of advanced knowledge-based computer systems to support scientific and technical creativity. Purpose of the work: development of services of personalized information support for scientific research in the institutions of NAS of Ukraine. The following services have been created: service for processing large-volume natural language texts, in particular, for highlighting terms, contexts of their usage and fragments of the hierarchy of terms; data sampling service for training distribution-semantic models; service for validation of publication metadata prepared for placement in the electronic library of scientific publications of Ukraine. Within the framework of these services the following options have been implemented: pipeline processing elements of Ukrainian texts, inter alia, the functions and procedures for extracting text data from electronic documents in docx, .pdf and .txt formats; normalization of texts (linguistic preprocessing); the function of automatic spell correction; elements of lexical analysis of the text. The service for validation of publication metadata can perform: collection of metadata in different schemes by unique metadata identifiers and time metrics, support basic and advanced search using field transitions from all archives that can be indexed. The system-ontological analysis of the formation of applications for an invention has been performed based on the ontological representation of their semantic content, as well as the appropriate processes have been developed, which are described by ontological models of the objects, processes and tasks of the problem space. Regarding the presentation of design knowledge, the following are proposed: weak methods of solving problems in the absence of experience and accurate mathematical models; strong methods of solving problems with existing experience in designing similar samples or accurate mathematical models together with quality criteria for results

Product Description

popup.authors

Velychko Vitalii Yu
Kurgayev Oleksandr P
Malachov Kyrylo S
Novytskyi Oleksandr V.
Petrenko Mykola G.
Stryjak Oleksandr E

popup.nrat_date

2021-01-11