【摘要】 范例推理(CBR)是一种用先前求解问题的经验和方法,通过类比和联想来解决当前相似问题的推理技术,它是动态决策环境下求解不良结构问题的常用方法。GIS系统作为一种新兴的地学工具,具有很强的空间分析能力,但由于地学问题的复杂性,一些地学现象很难用确切的模型进行模拟和预测。考虑到范例推理系统在处理半结构化和非结构化问题方面的出色能力,文中探讨了一个基于范例推理的GIS系统结构,并给出了地理范例的构建方法和表达模型。更多还原

【Abstract】 Cass-based reasoning system is a very important machine-learning method.To explain a new object case,CBR system will search original case in the case-base,find an old case which is the most similar to the new object base.GIS,as a new geography tool,is famous for its powerful space analysis ability.But because of the complexity of the geography problems,it is difficult to construct an assured model to predict geography phenomena.Considering the fineness ability to solve the half-structured and non-structured problem of the CBR system,the paper discussed the structure of the CBR-based GIS system.It presents how to construct and how to explain a geo-case.更多还原