Abstract:
The paper describes the methodology for automation of normative profiles forming in software certification. It is noted that currently, the practical evaluation for software certification at the stage of normative profiles forming is largely a manual analysis of large volumes of normative and project documentation submitted by texts in natural language, which leads to certain subjectivity of expert assessments, reduce their completeness and reliability, and the use of standard methods is ineffective due to their universality. The degree of automation of the certification procedure, in general, and normative profiles forming, in particular, are analyzed. The structure of the regulatory profile, the types of regulatory profiles, as well as options for its formation are given. The typical errors that can occur when automating the procedure of normative profiles forming are discussed. It is proved to use the ontological environment that in order to automate the process of normative profiles forming. A model of the compositional structure of the standard text is presented. The use of the semantic integrity kernel for the user request reflecting the relationship between subject and predicate lexical units is proposed. The description of the subject area of software certification is presented in the form of an extended ontological model, which includes the ontology of software quality criteria, the ontology of the standard database of software certification, the ontology of the core of semantic integrity, the ontological system of the organization of output on knowledge. The results of the research can be used in the development of an intelligent decision-making dialogue system for the auditor of the certification center in order to improve the efficiency of the auditor's work by automating the routine process, as well as reducing the risk of making the wrong decision due to insufficient qualification of the person making the decision.
Fulltext:
PDF file (641 kB)
