Case-based Maintenance Model for handling Relevant and Irrelevant Cases in Case-based Reasoning System

Abstract

Case-based maintenance can be resource intensive and requires significant time and effort to collect and analyse all cases. This can lead to inefficiencies and high costs in the entire case-based reasoning system. Accordingly, the Relative Coverage Condensed Nearest Neighbour had been created to reduce the number of cases in a dataset by selecting a subset of representative cases, whereas maintaining the overall performance of the whole system. Besides, Footprint utility deletion is a type of case deletion algorithm that can remove redundant or irrelevant cases from a storage, though maintaining the system’s competency. Recently, Hybrid approach was given to ensure that the case-base remains up-to-date and relevant, while also reducing its size and complexity. However, the results from using these approaches seem to be improved for the better performance. Therefore, the proposed model is developed, which comprises two main phrases by using case-based reasoning and identifying relevant and irrelevant cases to provide better results. The reduction size of case-base is lower than the traditional studies approximately 1-9% and also gives higher percentage of solving problems about 1-7%, while the average problem-solving time is shorter than them nearly at most 8 times.