Data warehouse model construction system and method based on multi-dimensional theory

Abstract

The invention discloses a data warehouse model construction system based on a multi-dimensional theory and a construction method thereof, and belongs to the field of data warehouses. The system and the method are improved on the basis of the multi-dimensional model design, and fully utilizes metadata to automatically construct a star-shaped multi-dimensional data model and shorten the data model design by precipitating the metadata of a design model in a meta-model. The system comprises a metadata model module, a dimension and hierarchy module, a measurement group and measurement module, a data cube module, a physical model module and the like. Automatic generation of a data warehouse physical model is realized by constructing a series of modules. The multi-dimensional design meta-model isindependently constructed, a traditional data model design tool is abandoned, automation of physical model scripts based on data model metadata is achieved, compared with traditional design, the datamodel design process is shortened, and the data warehouse construction efficiency is improved.

Description

technical field [0001] The invention belongs to the field of data warehouses, and in particular relates to a data warehouse model construction system and a construction method based on multidimensional theory. Background technique [0002] In the process of building a data warehouse, the data warehouse model is the foundation, and the data processing process (ETL) is the core. [0003] Generally, the traditional construction process is to first construct a multidimensional data model through data modeling tools. In the process of building the model, a large number of dimension tables and fact tables need to be manually created, which consumes a lot of manpower, and it is difficult to use standard specification constraints. , which leads to non-standard and inconsistent output data models, which brings unnecessary workload overhead and hidden dangers to subsequent data processing and data applications. [0004] The second is the data processing process (ETL). The traditional...

AI Technical Summary

Problems solved by technology

[0003] Generally, the traditional construction process is to first construct a multidimensional data model through data modeling tools. In the process of building the model, a large number of dimension tables and fact tables need to be manually created, which consumes a lot of manpower, and it is difficult to use standard specification constraints. , which leads to non-standard and inconsistent output data models, which brings unnecessary workload overhead and hidden dangers to subsequent data processing and data applications

[0004] The second is the data processing process (ETL). The traditional ETL process is to manually design the ETL mapping processing relationship according to the data warehouse model, and then the ETL engineer develops the ETL task according to the ETL design document. It is difficult to ensure that the ETL task in this way is consistent with the data warehouse model. The design is consistent, which brings a lot of testing and iterative revision workload

Images