Application of Machine Learning Techniques for Reducing Retrieval Time in Large Case-based Reasoning Systems

F. Martinez and L. Garrido (Mexico)


case-based reasoning, machine learning, case retrieval, clustering, decision tree.


In real environments case bases are usually very large, so it is often a slow process to obtain the right case that will work for a particular problem solving situation using case based reasoning (CBR). It is important to reduce retrieval time in CBR systems, in order to give better response time to the final user. In this paper, we present a novel approach to reduce the retrieval time in CBR systems. The approach that we show in this paper is a combination of cluster and decision tree techniques. This combination makes possi ble to build indexing structures in an automatic way. Our experimental results, based on two public domain datasets, show that employing our new approach improves retrieval time in CBR systems without losing significant accuracy degree in large case bases.

Important Links:

Go Back