Creating SQL Query to determine similar customers

Question

I have a table that contains two columns: CustomerID and ItemID.
There are 480 CustomerIDs and more than 5000 ItemIDs. The table shows what all items have been bought by each customer.
I need to write a procedure where i can determine which all customers are 'Like Customers' (It means if a Customer C1 buys Items A, B and C where another customer C2 buys A, B, C, D, E then we can say that C1 and C2 are 'Like Customers' since, they both
bought A, B, C.  It is also possible that both the customers bought B,C,D or C,D,E or any other combination)
I would also like to know whether this kind of problem can be better solved in SSMS or SQL Server Business Intelligence Devp Studio (using data mining techniques)?

Answer 1

For future reference, posting DDL and providing sample data makes it easier to answer your questions...
Anyway, based on the size of your table, simple T-SQL should do the trick.
-- a little test data.
CREATE TABLE #temp (
CustomerID INT,
ProductID INT
INSERT #temp (CustomerID,ProductID) VALUES
(1,1),
(1,2),
(1,3),
(1,4),
(1,5),
(2,4),
(2,5),
(2,6),
(2,7),
(3,2),
(3,3),
(3,8),
(3,9),
(4,10),
(4,11),
(4,12),
(4,13),
(5,1),
(5,2),
(5,12),
(5,13),
(6,1),
(6,5),
(6,7),
(6,10),
(7,1),
(7,2),
(7,3),
(7,4),
(7,5)
-- the actual solution
SELECT
t1.CustomerID AS Customer1,
t2.CustomerID AS Customer2,
COUNT(*) AS CommonItemCount
FROM
#temp t1
JOIN #temp t2
ON t1.CustomerID < t2.CustomerID
AND t1.ProductID = t2.ProductID
GROUP BY
t1.CustomerID,
t2.CustomerID
HAVING
COUNT(*) >= 2 -- set this to whatever threashold value you deem as "similar"
ORDER BY
t1.CustomerID,
t2.CustomerID
HTH,
Jason
Jason Long