摘要
基于路径映射的XML数据集成系统在查询重写时可能会生成不合理的子查询 .为了让生成的各个子查询中的实体属性是一致的 ,按照模式之间的路径映射提出了映射依赖的概念 ,并设计了一种查询重写的方法 .重写时依次遍历查询树的各个结点 ,记录每个中间结果的PC环境 ,根据启发式规则判断PC环境与当前映射的依赖是否保持一致 .
In order to make attributes of entities consistent in every subquery, a concept of mapping dependence is proposed in terms of the path mapping between schemata, and a method of reformulation is given. By this method, when traversing each node of query tree in reformulation, the system holds a PC-context for every temporary result reformulated, and in terms of heuristic rules judges whether or not the PC-context and the dependence of current mapping are consistent. The time complexity of algorithm in this method is linear with the size of data sources.
出处
《东南大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2004年第4期441-445,共5页
Journal of Southeast University:Natural Science Edition