Exploiting past experience - Case-based decision support for soccer agents

Abstract

Selecting and initiating an appropriate (possibly cooperative) behavior in a given context is one of the most important and difficult tasks for soccer playing robots or software agents. Of course, this applies to other complex robot environments as well. In this paper we present a methodology for using Case Based Reasoning techniques for this challenging problem. We will show a complete workflow from case-acquisition up to case-base maintenance. Our system uses several techniques for optimizing the case base and the retrieval step in order to be efficient enough to use it in a realtime environment. The framework we propose could successfully be tested within the robot soccer domain where it was able to select and initiate complex game plays by using experience from previous situations. Due to space constraints we can give just a very brief overview about the most important aspects of our system here. © Springer-Verlag Berlin Heidelberg 2007.