Data Analytics: Exploiting the Data Warehouse - Data Warehouse Systems: Design and Implementation - page 354

Fig. 9.12 Itemsets for the Northwind case study
The query returns the following result:
MINING PARAMETERS
MAXIMUM ITEMSET COUNT=200000, MAXIMUM ITEMSET SIZE=3,
MAXIMUM SUPPORT=1, MINIMUM SUPPORT=7.53768844221106E-03,
MINIMUM IMPORTANCE=-999999999, MINIMUM ITEMSET SIZE=0,
MINIMUM PROBABILITY=0.1
Our next query retrieves all the itemsets, together with a nested table (an
attribute that is actually a table) that lists the products included in each
itemset. The NODE NAME column contains the unique ID of the itemset
within the model. In this example, the nested table has been flattened (using
the FLATTENED keyword). Thus, an itemset will generate as many rows in
the result as products such itemset contains. This is the case, for instance, of
nodes 74 and 75 , which contain two items:
SELECT FLATTENED NODE NAME AS [Node],
NODE PROBABILITY AS [Probability],
NODE SUPPORT AS [Support],
(SELECT ATTRIBUTE NAME AS [Name]
FROM NODE DISTRIBUTION) AS Products
FROM Association.CONTENT
WHERE NODE TYPE = 7
ORDER BY NODE NAME
Each itemset is contained in its own node (as indicated by NODE TYPE =
7 ). This query returns the following table: