范例推理在口腔创伤诊断系统中的应用

Application of Case-based Reasoning in Oral Wound Diagnosis

鉴于传统口腔创伤诊断专家系统在知识获取上存在的瓶颈问题,提出一种基于范例推理(CBR)的口腔创伤诊断系统。首先,针对传统口腔创伤诊断的思维推理过程,给出了口腔创伤诊断问题中各要素的系统化定义,建立系统的的概念模型。然后,具体讨论了口腔创伤诊断系统中范例库和结果库的建立,并将其用关系数据库的格式存储。最后,应用模糊集和灰色关联理论,给出了关联度的定义和计算方法,从而能够准确地检索到相似诊断范例,进而对口腔创伤诊断给出定性或定量的分析结果。

To conquer the bottleneck problems existing in traditional Oral Wound Diagnosis expert system, a CBR （case-based reasoning） Oral Wound Diagnosis system is brought forward. Firstly, systematic definitions of several elements are proposed and a conceptual model is built for traditional thought reasoning of Oral Wound Diagnosis. Secondly, the construction of cases base and results base are discussed in detail and store them in the format of relation database. At last, definitions and calculation methods of relational degree are given using fuzzy sets and Grey-relational theory. Accordingly, similar diagnosis cases can be retrieved concisely so that qualitative and quantitive analysis results are given.