An Algorithm About Cube For OLAP Query Based On Partition

With the data cube playing an important role in the OLAP,and the exploded of data quantity and dimensions,the traditional algorithms cannot satisfy general requirements.The traditional data cube algorithm is not so efficiency in the big data and high dimension's situation,it's cost will be unacceptable,and there are more and more problems exposed.Data cube's space cost is the bottle neck of the OLAP,in order to solve these disadvantages,there are more and more algorithms focus on this problem,and they all earn a nice performance,but now with the exposed of the data quantity,and the complex conditions in the reality,the limit of the algorithm is disclosed.So this paper is devoted to solve these questions,and it proposes a new data model based on partitions.This concepts' kernel is to inverse the data,it means the data will from attributes to IDs.This model can link different attributes' data to replace the traditionals way.It can get the aggregate result from the customize data structure in query.Then it's realized by a data warehousing with the specific application,and can get the final result and parameters.The algorithm has been achieved by the data warehousing system we established.It could be find that the space efficiency about this algorithm is better than shell-fragment data cube algorithms.Besides,it's also competitive in temporal.So it obtains an efficitive algorithm by the partition concept.