Improving the Efficiency of Solution Search Systems Based on Precedents

Abstract

In this paper, actual issues of improving the efficiency of solution search systems based on precedents -Case-Based Reasoning Systems (CBR systems) are considered. To improve the efficiency of CBR systems and accelerate the search for solutions, it is proposed to use a modified CBR cycle, which allows to create a base of successful and unsuccessful precedents and reducing the number of precedents in the database of successful and unsuccessful precedents through the use of classification and clustering methods.