Case-based reasoning (CBR) in development control

Abstract

Case-based reasoning (CBR) is a type of knowledge-based systems (KBS) whose philosophy is to use previous cases to interpret or solve a new problem. There are many advantages of CBR over rule-based reasoning which is the method adopted by most current KBS for development control. The knowledge representation scheme of CBR is better than rule-based reasoning in recording and representing knowledge that is hard to express with explicit rules or is too case-specific. It can overcome the problems with black box inference process of rule-based reasoning. CBR can simulate the present working style of a planner in dealing with development applications which is based on his/her knowledge of past application records. The results derived directly from real cases in CBR are more convincing and acceptable to planners. A system that integrates a CBR shell (ESTEEM) and GIS package (ArcView) was developed to demonstrate the process in building a case-based system (CBS) to assist planners in the Planning Department of Hong Kong in handling applications in development control. Because it is case-based and not rule-based, the system is also applicable to other cities in dealing with development applications.