The object of research is the process of automated construction and enrichment of a knowledge base. An approach has been defined to address the problem of automated knowledge base construction in information system development projects. Research has been conducted on models and methods for the automatic construction of a knowledge base for process management information systems. It has been determined that finite predicates, as well as predicate models describing IT project artifacts, are suitable as the mathematical apparatus. The method for automated construction and enrichment of a knowledge base for process management information systems has been improved based on modified predicate models. The modification involves adding new predicates and aligning event logs with classes of artifacts and their properties. The feasibility of applying the proposed method in decision support tasks under intrusion detection conditions for processes in computer systems has been verified. The architecture and description of the main characteristics of the automated knowledge base construction system for decision-making tasks in IT enterprises, are provided. The proposed solutions have been tested during the automated construction of a knowledge base in an IT project. Advantages and disadvantages of the method were identified during the automated construction of a production knowledge base. Further development of the modified method may involve analyzing the success of predictions made using the knowledge base and adjusting the initial set of predicates based on the results of this analysis.
Read full abstract