The Research Of Data Warehouse Model Based On Theory Of Quotient Space

In the knowledge-based economy era of information technology highly be developed, on the one hand, the amount of data in various practical applications is rapidly expanding; on the other hand, intense competition caused businesses which stand in the era of knowledge economy to require the analysis point of decision support system, not only comprehensiveness in the breadth of information, and requirements the level of granularity in the depth of information. As the data warehouse is designed for on-line analysis processing, used to save for inquiry and decision-making, historical and integrative data. Decision support system based On-line analysis processing Presents the challenge on ability of the data warehouse model effectively managing massive data. In view of this, building the multidimensional data models which manage effectively massive data and support for complex hierarchical structure is becoming new hot spots of data warehouse technology research. This paper, under the guidance of theory of quotient space, taking massive data's effective hierarchical management as the goal, explores to proposing multi-dimensional data warehouse model based on the theory of quotient space. The model establishes one to one mappings between equivalence relations of theory of quotient space and dimension hierarchical attribute of model, through cursor which is named degree of coarse of equivalent relations modulating granular transformation of dimension hierarchy attributes , realizes data warehouse core operations such as aggregated operation, and through the way of the multi-dimensional array combining tree index realizes to store effectively materializing cube of the multi-dimensional data model , this not only maintains relational condition of original data semantics, but also increases Index way for OLAP according to the principle of false security of the theory of quotient space. The whole model's choice takes goal to enhance the OLAP inquiring speed, to optimize the OLAP queries, finally, with the data and the performance of Teaching Quality Scrutiny Data Warehouse, the choice of model is operating to the test. The main contributions of this thesis are as follows:1. A multi-dimensional data model of data warehouses, which effectively describing complex hierarchical dimension structure, is proposed. Based on flexibility transformation technology with size of the coarse granular level of quotient space granularity theory, this model not only supports basic algebraic operation of data set, but has the superiority in aggregation operation of nimble size of coarse granular transformation between levels of dimension ,(such as aggregation operation of levels, aggregation operation of dimensions ), and aggregation operation is basic step in the OLAP core operation (for example operation of roll-up, drill-down and so on) , has important status in the OLAP elementary operation and the OLAP inquiry.2. Aggregation cube storage Implementation of the multidimensional hierarchical data model is explored. With characteristics of the hierarchy structure which dimension levels of Multidimensional Data Model have, it is considered that materializing cubes realize storage by method of building the dimension hierarchy tree. Moreover corresponding algorithm named HDEKC, which to realize storage functionality, is proposed.3. It is researched that using index method of quotient space theory to realize the power aggregated operation and OLAP inquiry operation of multidimensional data model. A algorithm named AQCA which realizes the aggregated operation of Data Cube under theory of quotient space is proposed.