一种专家系统知识获取时的属性约简算法

知识获取是构造专家系统的“瓶颈”，提供准确的推理知识是进行决策规划的关键。文中运用粗糙集理论，通过粗糙集的约简消除冗余的条件属性，实现对知识库的精简。首先研究知识获取，在阐明知识的层次结构基础上，给出了概念化、形式化、知识库求精三个知识获取过程；然后研究属性约简算法，在研究集合差异度和属性的重要性、约简算法推导过程的基础上，给出了属性约简算法的六个步骤。最后根据属性约简算法及其步骤，对功能点分析法构建软件成本估算专家系统时，组成技术复杂因子的14个因素进行了约简。

An Attributes Reduction Algorithm of Expert System Knowledge Acquisition

Knowledge acquisition is the ＂bottleneck＂ of construction expert system, to provide an accurate inference of knowledge is the key decision-making plan. It uses the rough sets theory, eliminate redundant condition attribute through the rough sets reduction to achieve the streamlining of the knowledge library. First study the knowledge acquisition, in exposition knowledge hierarchical structure foundation, has given three knowledge acquisition of the conceptualization, formal, the knowledge library refinement and so on. And then study attributes reduction algorithms, on the basis of researching sets difference and the attribute importance, the reduction algorithms inferential reasoning process, has given the attribute reduction algorithms six steps. Finally, according to the attributes reduction algorithms and the steps, 14 factors of the composition technology complexity factor are reduced when software cost estimation expert system is constructed by function analysis method.