Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»

There can be any number of foreign key relationships to the dimension tables. A foreign key relationship is created when there is a need to examine the foreign key data along with data in the fact table.

One of the interesting aspects of the star join is that in many cases textual data is divided from numeric data. Consider the diagram in Figure 3.54. Textual data often ends up in the dimension tables, and numeric data ends up in the fact table. Such a division occurs in almost every case.

The benefit of creating star joins is to streamline data for DSS processing. By prejoining data and by creating selective redundancy, the designer greatly




order    product

Figure 3.52 A three-dimensional perspective of the entities shows that the entities are anything but equals. Some contain far more occurrences of the data than others.

dimension    fact table    dimension

tables    tables


numeric    character

Figure 3.54 In many cases, the fact table is populated by numeric data and foreign keys, while the dimension table is populated by character data.

simplifies and streamlines data for access and analysis, which is exactly what is needed for the data mart. Note that if star joins were used outside of the DSS data mart environment, there would be many drawbacks. Outside the DSS data mart environment, where update occurs and where data relationships are managed up to the second, a star join most likely would be a very cumbersome structure to build and maintain. But because the data mart is a load-and-access environment, because the data mart contains historical data, and because massive amounts of data need to be managed, the star join data structure is ideal for the processing that occurs inside the star join.

Скачать в pdf «Building the Data Warehouse»