模糊概念图知识表示及其推理机制研究  

Research on knowledge representation and inference mechanisms about fuzzy conceptual graphs

作　　者：刘培奇[1] 李增智[2] 

机构地区：[1]西安建筑科技大学信息与控制工程学院,西安710055 [2]西安交通大学电子与信息工程学院,西安710049

出　　处：《计算机应用研究》2010年第6期2119-2122,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(60673170);陕西省教育厅自然科学基金专项资助项目(08JK318);校人才基金资助项目(RC0618)

摘　　要：通过对现有模糊概念图的研究,针对概念的所指域与模糊信息间的冗余问题和用模糊度表示模糊概念问题,提出一种改进的模糊概念图知识表示方法。在改进的模糊概念图中,用模糊集合表示概念图中的模糊概念和模糊关系,并将模糊概念的所指域同模糊集合合并,减少信息冗余。根据改进的模糊概念图,重点研究了模糊概念图的匹配推理机制,设计了基于语义约束的匹配推理算法,并定量分析了算法的时间复杂度和空间复杂度。经过在《计算机文化基础》课程中实验测试,算法反映了考生主观题的答卷情况,同人工阅卷结果基本一致。By analyzing the existing fuzzy conceptual graphs, detected the information redundancy between referent and fuzzy degree of the concept. At the same time, fuzzy degree of some concepts doesn’t reflect fuzzy uncertainty of the concept completely. Based on analysis, put forward a new fuzzy conceptual graph. In the improved fuzzy conceptual graphs, used the fuzzy set to replace the fuzzy degree and enlarge the referent. It reduced information redundancy and increased the ability of the know-ledge representation. According to the new fuzzy conceptual graphs, studied the matching reasoning mechanism of the fuzzy conceptual graphs, and designed the matching reasoning algorithms based on semantic constraint. The paper accomplished the quantitative analysis of the time complexity and space complexity of the algorithm. Via the experimental test in the culture foundation of computer science, the algorithm is the same as manuals of checking over subjective examination basically.

关 键 词：模糊概念图 知识表示 知识推理 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]