Overview of Data Mining - Java Data Mining: Strategy, Standard, and Practice - page 7

SELECT store.name, sum (sale_amount) AS sales_sum
FROM store, sales, date, product
WHERE product.name 'Portable DVD Player' AND
product.id sales.product_id AND
sales.store_id store.id AND
sales.date date.date AND
date.quarter '2Q06'
GROUP BY store.id
ORDER BY sales_sum

Figure 1-1 shows four tables: SALES, PRODUCT, STORE, and
DATE. The SALES table contains the sales amount and keys into the
other three tables, indicating the product sold, the store that sold it,
and the date that it was sold. The PRODUCT table provides the name
of the product as well as a hierarchical grouping of that product into
higher-level categories. A product belongs to a particular group,
which belongs to a more general category (e.g., Coca-Cola 20-ounce
soda may have a product group of “colas” and be in the product
category of “soft drinks”).
Another technology to distinguish from data mining is Online
Analytical Processing (OLAP) technology [Wikipedia-OLAP 2006].
OLAP supports summaries, forecasts, and trend analysis. For example,
summaries often involve “rolling up” data to different levels of
granularity as defined by a dimension, such as Date, which can

DATE STORE

id
name
city
state
region
country

date
week
month
quarter
year

SALES

sale_amount
date
product_id
store_id

PRODUCT

id
name
product_group
product_category

Figure 1-1 Example of a star schema.