摘要
描述怎样从关系数据库中根据各种约束条件得到EER(extend entity—relationship)图进而解决对关系数据库的语义理解问题。基本思想是基于主键及外键约束来建立初始关联,包括所有的一元、二元及多元关系。然后确定基数约束,进而优化消除对称、传递等关系,并分辨出哪些是is-a关系,哪些是part-of关系及关联关系,并通过集成,将表达相同意思的同类实体合并。由这个过程就可以抽取出关系数据库的语义信息。
In order to resolve the problem of semantic understand, we get EER(extend entity--relationship)graph bases on the dependency relation of database. The main idea is found out the CandidateKey and the ForeignKey to build the initial graph, include all the Unary relationship,binary relationships and N - ary relationship. After make sure the cardinalities, we need to delete the symmetric and transitive relationship. Farther, analyses these relationships to find out which are is- a relationships and which are part- of relationships, in order to integration use some rule combine with the class which expression the same mean. At last we could get the semantic information from relationship database.
出处
《计算技术与自动化》
2007年第3期104-109,共6页
Computing Technology and Automation
关键词
关系数据库
候选键
外键
EER图形
语义
relationship database
candidatekey
foreignkey
EER graph
semantic