Association Rule Model Nugget Settings

This Settings tab is used to specify scoring options for association models (Apriori and CARMA). This tab is available only after the model nugget has been added to a stream for purposes of scoring.

Maximum number of predictions Specify the maximum number of predictions included for each set of basket items. This option is used in conjunction with Rule Criterion below to produce the “top” predictions, where top indicates the highest level of confidence, support, lift, and so on, as specified below.

Rule Criterion Select the measure used to determine the strength of rules. Rules are sorted by the strength of criteria selected here in order to return the top predictions for an item set. Available criteria are shown in the following list.

• Confidence
• Support
• Rule support (Support * Confidence)
• Lift
• Deployability

Allow repeat predictions Select to include multiple rules with the same consequent when scoring. For example, selecting this option allows the following rules to be scored:

bread & cheese -> wine
cheese & fruit -> wine

Turn off this option to exclude repeat predictions when scoring.

Ignore unmatched basket items Select to ignore the presence of additional items in the item set. For example, when this option is selected for a basket that contains [tent & sleeping bag & kettle], the rule tent & sleeping bag -> gas_stove will apply despite the extra item (kettle) present in the basket.

There may be some circumstances where extra items should be excluded. For example, it is likely that someone who purchases a tent, sleeping bag, and kettle may already have a gas stove, indicated by the presence of the kettle. In other words, a gas stove may not be the best prediction. In such cases, you should deselect Ignore unmatched basket items to ensure that rule antecedents exactly match the contents of a basket. By default, unmatched items are ignored.

Check that predictions are not in basket. Select to ensure that consequents are not also present in the basket. For example, if the purpose of scoring is to make a home furniture product recommendation, then it is unlikely that a basket that already contains a dining room table will be likely to purchase another one. In such a case, you should select this option. On the other hand, if products are perishable or disposable (such as cheese, baby formula, or tissue), then rules where the consequent is already present in the basket may be of value. In the latter case, the most useful option might be Do not check basket for predictions below.

Check that predictions are in basket Select this option to ensure that consequents are also present in the basket. This approach is useful when you are attempting to gain insight into existing customers or transactions. For example, you may want to identify rules with the highest lift and then explore which customers fit these rules.

Do not check basket for predictions Select to include all rules when scoring, regardless of the presence or absence of consequents in the basket.

Generate SQL for this model When using data from a database, SQL code can be pushed back to the database for execution, providing superior performance for many operations.

Select one of the following options to specify how SQL generation is performed.

• Default: Score using Server Scoring Adapter (if installed) otherwise in process If connected to a database with a scoring adapter installed, generates SQL using the scoring adapter and associated user defined functions (UDF) and scores your model within the database. When no scoring adapter is available, this option fetches your data back from the database and scores it in SPSS® Modeler.
• Score outside of the Database If selected, this option fetches your data back from the database and scores it in SPSS Modeler.