一种多特征方查询的有效算法

Efficient algorithm of multi-feature cubes query

多特征方用于计算复杂数据挖掘查询,具有多粒度上多个依赖聚集计算的典型特点。现有的立方体粒度计算技术可以有效计算分布和代数多特征方,整体多特征方还没有提出有效的计算方法。在分析整体多特征方查询特点的基础上提出了一种优化算法:先将立方体分块,再采用冰山查询技术动态选择数据,最后采用查询结果重用技术。实验表明,该优化算法能有效提高整体多特征方查询的性能。

Multi-feature cubes （MF-cubes） compute complex data mining queries involving multiple dependent aggregates at multiple granularities. Currently, most proposed cube computing algorithms efficiently process the distributive and algebraic MF-cubes. Yet, there is few algorithm for computing holistic MF-cubes. An algorithm is proposed to compute the Holistic MF-cubes, which combined three efficient techniques： Partition of data cube into smaller sub-cubes; A dynamic subset data selection strategy-iceberg query; Reuse the overlap result of previous queries. The experimental results show that the algorithm is highly efficient to the holistic MF-cubes queries.