作者: Kathleen Hanney , Mark Keane , Barry Smyth , Padraig Cunningham
关键词: Artificial intelligence 、 Multiple case 、 Machine learning 、 Computer science 、 Data mining 、 System development
摘要: This paper shows that the use of adaptation knowledge in CBR systems is heavily dependent on certain task and system constraints. Furthermore, type used performing specific tasks quite regular predictable. These conclusions are reached by reviewing forty-two classifying them according to three taxonomies: an adaptation-relevant taxonomy systems, a knowledge. We then show how different cluster with respect interactions between these taxonomies. The designer may find partition division suggested this useful. Moreover, help focus initial stages development suggesting (on basis existing work) what types should be supported new system. In addition, provides framework for preliminary evaluation comparision systems.