摘要
提出了一种基于模糊聚类的属性匹配算法。该算法采用能综合反映属性名称相似性和语义相似性的模糊相似关系,提高了属性匹配的准确率;以等价闭包法对相似属性进行模糊聚类,得到多层次属性分类结果,更客观真实地反映了属性匹配的模糊性;同时,属性匹配过程中不需要设置匹配参数,避免了人为造成的误差。
By means of the fuzzy similarity of attribute, the attribute matching algorithm based on fuzzy clustering improves the veracity of attribute matching. It uses the equivalent closure to cluster the attributes fuzzily, and gets the multilevel attributes classified results. During the matching of the similar attributes with the fuzzy clustering, it doesn't need to set the parameter of clustering, and avoids the manual errors.
出处
《模糊系统与数学》
CSCD
北大核心
2006年第6期96-102,共7页
Fuzzy Systems and Mathematics
基金
国家自然科学基金资助项目(60172012)
关键词
模糊聚类
属性匹配
数据集成
Fuzzy Clustering
Attribute Matching
Data Integration