非一致性数据库的多关系聚集查询重写

Multi-Relations Aggregation Query Rewritings in Inconsistent Databases

基于非一致性关系数据库的选择连接查询技术,提出了基于非一致性数据库多个关系上的聚集查询重写方法。该聚集查询重写方法先通过查询出多关系上的一致性结果,然后进行分组聚集,返回聚集表达范围边界值。实验采用TPC-H决策支持基准进行性能研究,结果表明重写查询比初始查询的执行时间要长,但还是可以接受的,因此该方法是有效的。

Based on the technique of selective join query on inconsistent datahases, proposes an approach, which can overcome the obstacles in the field of multi- relations aggregation query rewritings in inconsistent database. Can search the consistent results on multi-relations and duster the results by means of grouping aggregation to return the extremurn in the range of aggregation values. The experiment use TPC-H to study the capability. The result demonstrates that the rewriting query have a longer carrying-out time than the original query, but it＇s acceptable,so the method is operative.