Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»


acceptable reject rate expediter last order date last order amount last delivery to shipping manifest order amount


part-id description order unit lead time


acceptable reject rate shipping manifest


part-id


p art- id qty on hand last order date last order amount last delivery to order amount



primary substitute safety stock primary supplier expediter


Figure 3.9 An example of stability analysis.


result of stability analysis (which usually is the last step of data modeling before physical database design) is to create groups of data with similar characteristics.


There is, then, a common genesis of data models. As an analogy, say the corporate data model is Adam, the operational data model is Cain, and the data


warehouse data model is Abel. They are all from the same lineage, but at the same time, they are all different.


The Data Warehouse Data Model


(Other books have been written on data modeling, detailing several different approaches. Any number can be used successfully in building a data warehouse. The approach summarized here can be further explored in my previous book Information Systems Architecture: Development in the 90s [QED/Wiley, 1992].)


There are three levels of data modeling: high-level modeling (called the ERD, entity relationship level), midlevel modeling (called the data item set, or DIS), and low-level modeling (called the physical model).


The high level of modeling features entities and relationships, as shown in Figure 3.10. The name of the entity is surrounded by an oval. Relationships among entities are depicted with arrows. The direction and number of the arrowheads indicate the cardinality of the relationship, and only direct relationships are indicated. In doing so, transitive dependencies are minimized.

Скачать в pdf «Building the Data Warehouse»