Knowledge Maintenance of Case-Based Reasoning Systems

The SIAM Methodology

Roth-Berghofer, Th.R.
Pub. date
March 2003
262 of Dissertations in Artificial Intelligence
ISBN print
Artificial Intelligence, Computer & Communication Sciences, Computer Science

For all technical systems that are widely used, maintenance processes are available. Now that Case-Based Reasoning (CBR) has been used more and more over the last decade, a maintenance methodology was overdue. Thus, it is not only of academic but also of commercial interest to provide structured and cost effective maintenance of deployed CBR systems. The main idea to solve the maintenance problem is to employ a control loop on CBR systems, to observe its performance and to counteract any change that is not desired. Those changes, in principle, come from two sources. Changes may be due to the use of the system, for instance, when it learns new cases, or it may be due to changes of the environment such as changes in legislation, in technology, or in norms. In order to realize the control loop, the most widely used CBR process model was enhance and placed in a broader context that regards the whole life cycle of a CBR system. The four phases Setup, Initialization, Application, and Maintenance form the basis of the SIAM methodology. SIAM extends the INRECA methodology that provides a framework for developing CBR solutions. With SIAM, better maintainable Case-Based Reasoning applications can be developed.