Abstract: Disclosed is a multi-dimensional decomposition computing method and system, related to the field of multi-dimensional decomposition computing technique, applicable to performing multi-dimensional decomposition on big data to reduce computation complexity. The method executes: generating an recursion topology based on pre-processed big data; the recursion topology comprising dimension combination and recursion path among the dimension combinations; based on default fixed strategy, defining a fixed dimension combination and defining a computation path forming an optimized fixed dimension combination; based on the recursion topology, generating computation tasks; and based on the fixed dimensional combination and the computation path of the optimized fixed dimension combination, activating the computation tasks, computing the computation tasks and obtaining computing results. The disclosed solution is applicable to multi-dimensional decomposition.

Abstract: A multidimensional data process method and device thereof are disclosed in the present embodiments. The method comprises: obtaining attribute information in dimension information and hierarchical relationship information of respective attributes in a data service, and generating a recursive topology; and the recursive topology comprises an attribute composition set and recursive paths of respective attribute compositions in the attribute composition set; confirming attribute compositions for querying in the attribute composition set; confirming attribute compositions for fixing in the attribute composition set according to the attribute compositions for querying; fixing the attribute compositions for fixing to form fixed attribute compositions and obtaining index data in accordance with the fixed attribute compositions; recursing index data in accordance with the attribute compositions for querying. The present invention is applied to the multidimensional data analysis.

Abstract: Disclosed is a multi-dimensional decomposition computing method and system, related to the field of multi-dimensional decomposition computing technique, applicable to performing multi-dimensional decomposition on big data to reduce computation complexity. The method executes: generating an recursion topology based on pre-processed big data; the recursion topology comprising dimension combination and recursion path among the dimension combinations; based on default fixed strategy, defining a fixed dimension combination and defining a computation path forming an optimized fixed dimension combination; based on the recursion topology, generating computation tasks; and based on the fixed dimensional combination and the computation path of the optimized fixed dimension combination, activating the computation tasks, computing the computation tasks and obtaining computing results. The disclosed solution is applicable to multi-dimensional decomposition.

Abstract: A multidimensional data process method and device thereof are disclosed in the present embodiments. The method comprises: obtaining attribute information in dimension information and hierarchical relationship information of respective attributes in a data service, and generating a recursive topology; and the recursive topology comprises an attribute composition set and recursive paths of respective attribute compositions in the attribute composition set; confirming attribute compositions for querying in the attribute composition set; confirming attribute compositions for fixing in the attribute composition set according to the attribute compositions for querying; fixing the attribute compositions for fixing to form fixed attribute compositions and obtaining index data in accordance with the fixed attribute compositions; recursing index data in accordance with the attribute compositions for querying. The present invention is applied to the multidimensional data analysis.