In modern conditions of standardization, new tasks arise, which involve the transition of international and regional standards to the digital format of "smart" standards that can be used as full-fedged guidelines for industrial facilities. The introdu...
In modern conditions of standardization, new tasks arise, which involve the transition of international and regional standards to the digital format of "smart" standards that can be used as full-fedged guidelines for industrial facilities. The introduction of artifcial intelligence and machine learning in the standardization process will increase the level of automation and allow analyzing data presented in the form of a database, which contributes to better compatibility between systems and equipment of diferent manufacturers. This article presents a method of ontological design, which automates partial methodological examination of documents in accordance with the National Standardization Plan. As part of the experiment, an independent tool was developed from the existing repositories of regulatory documents. This approach is based on the extraction of qualitative data from textual information using an unsupervised automatic keyword extraction algorithm. The algorithm is based on statistical characteristics of local text extracted from individual documents. The advantage of the algorithm is that it does not require the creation of a learning corpus and can be applied to any text, and also has no limitations on the volume of processed data. The following components are presented in the article: model classifcation standards, entity extraction module, categorization module, thesaurus generation module and monitoring and analysis module. The calculation was carried out using formalization, defnition of the target function, establishment of the similarity measure, as well as description of the features of the document. The obtained results were tested, based on standards and rules in the feld of the aerospace industry.