论文部分内容阅读
传统的案例相似性度量方法无法很好地判断复杂案例库系统中某些特殊属性(如数值属性、模糊属性)的相似程度,从而限制了案例推理(CBR)系统在判断、评估、决策、推理方面的进一步应用.针对此问题提出混合概念格的方法,对案例库中可能涉及到的数字、符号、模糊属性等进行相似性度量.实验表明,该方法较好地解决了混合案例属性的相似性度量问题,提高了检索结果的查全率和查准率.
The traditional case similarity measurement method can not judge the similarity of some special attributes (such as numerical attribute and fuzzy attribute) in the complex case base system, which limits the advantages of CBR system in judgment, assessment, decision, reasoning This paper proposes a hybrid concept lattice approach to this problem and makes researches on the similarity of numbers, symbols and fuzzy attributes that may be involved in the case base.Experiments show that this method can better solve the similarity of the mixed case attributes The measurement of sexual problems, improve the search results of the recall rate and accuracy.