Knowledge discovery using maximization of the spread of influence in an expert system

The problem of knowledge acquisition (KA) and the automation of this process is a bottleneck of expert systems (ESs) and remains a topical issue. The present study is an attempt to develop an ES in which inference attributes are formed on the basis of automated discovery of tacit knowledge. The structure of the KA module provides an opportunity to discover new knowledge from the cognitive representations of domain models. The algorithmic processing mechanism consists of three blocks: cluster analysis, influence analysis, and modelling of the equivalent graph. In the proposed expert system based on effective controls, a method of finding influences that highlights the main characteristics of the system is presented. These parameters are used as attributes for the further development of the rule base, expertise, or independent use in decision‐making. A qualitative decrease in user subjectivity is achieved using this algorithm and through elimination of the knowledge engineer from the cycle of KA, which transforms it into a discovery knowledge cycle. Quality assurance of the knowledge base is achieved using three approaches: system presentation, mathematical support, and automation.