Semantic and Association Rule Mining-based Knowledge Extension for Reusable Medical Equipment Random Forest Rules

Abstract

In this article, we present a knowledge extension method for transforming and extracting Reusable Medical Equipment (RME) rule pattern sets from random forest models with the aid of the domain knowledge. Healthcare providers are increasingly relying on using Reusable Medical Equipment (RME) since it can be reused and reprocessed to multiple patients. Hence, estimating the maintenance (i.e., repair) cost during RME lifecycle has been a topic in healthcare domain. The aim of this research is to propose the method of knowledge extension based on the post-mining interpreted by RME ontology and statistical cost domain knowledge. The new RME ontology is to provide domain knowledge to help interpret the statistically supported strong patterns of pre-mined decision rules. This article applied the presented knowledge extension method to find the frequent rule patterns from the significant large volumes of decision rules of the non-profit hospital’s legacy database. Finally, the case study shows how specific rule patterns can be generated and the user can understand the patterns regarding cost with the statistical RME cost domain knowledge and RME ontology. The results show that the extension method can provide plausible reasoning for the interpreted patterns in terms of the specific scope information.