Multistep Decision-Making in Case-Based Planning

摘要：

During the course of decision-making in case-based planning, users experience and knowledge is very significant. This paper discusses how to incorporate them into PLANAUTHOR, a case-based planning system developed by us, before making a decision. Using templates, the authors give a new method, integrating case-based reasoning with decision tree algorithm, to make a multistep decision in case-based planning. The templates in this system were first used to obtain a plan frame in terms of notified situations. The integration helped the users decide whether a step should be carried out at the strategic level and how the step is performed at the tactical level, in turn. The results show that the integration can improve efficiency and quality of planning with templates. Most of all, past cases can make the users more confident about the solutions proposed by the system.

关键词： case-based reasoning planning with templates decision tree algorithm.

作者: Zhou Xi-Feng Shi Ze-lin Li Yi-fan Zhao Huai-ci Zhou Lei

作者单位: Shenyang Institute of Automation of the Chinese Academy of Science, Shenyang, China, 110016;Graduate Shenyang Institute of Automation of the Chinese Academy of Science, Shenyang, China, 110016

会议类型: 国际会议

会议名称: Firth IEEE International Conference on Cognitive Informatics(第五届认知信息国际会议)

会议地点: 北京

会议语种:英文

页码: 629-633

在线出版日期: 2006-07-17（万方平台首次上网日期，不代表论文的发表时间）