data mining : associate rules

i begin for Data mining ( analysis manager - sql server 2000)
i create some mining model with Microsoft_decision_ trees or Clustering. it's oK
but now i want to create a model to know "If customer 1 has product A in their basket, what products should I recommend ".

i read and see that "Microsoft analysis manager sp1" have support association rules.
i install pack 4. but i don't see anything else . how i create a model with association rules.

SQL Server 2005 provides the Microsoft Association Rules algorithm, which is veryu scalable and designed specifically for this kind of tasks.

However, you can still generate recommendations using SQL Server 2000 by using Decision Trees and a model with a nested table containing the products. An example of such model:

( CustomerID LONG KEY,

Products TABLE PREDICT

(

Product TEXT KEY

)

)

In the model, the Products nested table is used both as input and as output.

The prediction query will look like below:

select FLATTENED

TopCount( Predict([Products], INCLUDE_STATISTICS, EXCLUSIVE), $ADJUSTEDPROBABILITY, 5) FROM MyModel PREDICTION JOIN

(select

(SELECT 'Beer' AS Product UNION

SELECT 'Coke' AS Product) AS Products) as t

ON MyMOdel.Products.Product= t.Products.Product

Please post here any questions you might have

|||thanks, i will try now.