Learning Case-Based Reasoning Solutions by Association Rules Approach

Abstract

Case-Based Reasoning (CBR) is amply used as a method to solve problems in many domains. It involves retrieval and use of past cases to solve a new one. However, using past cases to solve a new one by this methodology is weighing a great deal. In this paper, we present our approach for reducing the search space solution through association rules and focus only interesting cases that meet only these rules. The reduced search space is then used by CBR to compute a solution for the new problem. Through this work, we aim to propose an approach that combines Association Rules and CBR to improve searching solution for similar cases. Thereafter, we test our approach by using real-life datasets.