Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»


Suppose a manager asks for a report from the data warehouse for 1995. The report is generated, and the manager is pleased. In fact, the manager is so pleased that a similar report for 1990 is requested. Because the data warehouse carries historical information, such a request is not hard to accommodate. The report for 1990 is generated. Now the manager holds the two reports—one for 1995 and one for 1990—in his hands and declares that the reports are a disaster.


The data warehouse architect examines the reports and sees that the financial statement for 1995 shows $50 million in revenue, while the report for 1990 shows a value of $10,000 for the same category. The manager declares that there is no way that any account or category could have increased in value that much in five years’ time.


Before giving up, the data warehouse architect points out to the manager that there are other relevant factors that do not show up in the report. In 1990, there was a different source of data than in 1995. In 1990, the definition of a product was not the same as in 1995. In 1990, there were different marketing territories than in 1995. In 1990, there were different calculations, such as for depreciation, than in 1995. In addition were many different external considerations, such as a difference in inflation, taxation, economic forecasts, and so forth. Once the context of the reports is explained to the manager, the contents now appear to be quite acceptable.


In this simple but common example where the contents of data stand naked over time, the contents by themselves are quite inexplicable and unbelievable. When context is added to the contents of data over time, the contents and the context become quite enlightening.

Скачать в pdf «Building the Data Warehouse»