提出一种高效的整体多特征方查询算法。该算法首先将数据立方体水平分块成多个小数据集,然后将各子查询中的聚集函数分类,并对其中的分布和代数聚集函数使用分布聚集特性优化计算,使得整体多特征方查询可以局部使用分布多特征方查询的优化计算方法。实验结果证明该方法可以有效地提高整体多特征方查询的效率。
This Paper proposed a fast algorithm for Holistic MF-Cubes queries, which included partition datacube into smaller sub-cubes and classify aggregating functions and identify the distribute ones, which can use the coarse-finer granularities computing property. Experiment shows that this algorithm is highly efficient to the Holistic MF-Cubes query.