基于集合运算的数据立方体结构

提出一种新的数据立方体结构,通过索引和集合的交并运算来获得查询结果,特别是在进行区域查询时,避免了将区域分解为点后再依次进行点查询的方式,从而在保持较少的磁盘空间和较好的点查询响应速度的情况下,改善区域查询的性能;同时给出其生成和查询算法,并使用合成数据和实际数据进行了实验验证.

Data cube structure based on operation of sets

This paper proposed a new data cube structure to avoid the model that firstly decomposed the range to a lot of points and then performed the point queries one by one. By means of index and intersection and union operations of sets, it not only also had low disk storage cost and comparatively good performance of point queries, but also improved the performance of range query. Then gave the algorithms of computing the data cube and queries also. In the end, the results of experiments by using both the synthetic and real data sets.