Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»

The departmental environment—sometimes called the data mart level, the OLAP level, or the multidimensional DBMS level—contains information useful to the different parochial departments of a company. There is a marketing departmental database, an accounting departmental database, an actuarial

a simple example—a customer



atomic/data    dept/data mart

warehouse customers by month

J Jones 123 Main Street Credit — AA

JJ JJoonneess 1986-1987

456 High St Credit — B

customers since 1982 with acct balances > 5,000 and with credit ratings of B or higher


JJ JJoonneess 1987-1989

456 High St

CCrreeddiitt AA

J Jones 1989-pres

123 Main St Credit; — AA

What is J Jones credit rating right now?

What has been the credit history of J Jones?

Are we attracting more or fewer customers over time?

What trends are there for the customers we are analyzing?

Figure 1.11

The kinds of queries for which the different levels of data can be used.

departmental database, and so forth. The data warehouse is the source of all departmental data. While data in the data mart certainly relates to data found in the operational level or the data warehouse, the data found in the departmen-tal/data mart environment is fundamentally different from the data found in the data mart environment, because data mart data is denormalized, summarized, and shaped by the operating requirements of a single department.

Typical of data at the departmental/data mart level is a monthly customer file. In the file is a list of all customers by category. J Jones is tallied into this summary each month, along with many other customers. It is a stretch to consider the tallying of information to be redundant.

Скачать в pdf «Building the Data Warehouse»