Market basket data identifies the items sold in a set of baskets or transactions. Oracle Data Mining provides the association mining function for market basket analysis.
Association models use the Apriori algorithm to generate association rules that describe how items tend to be purchased in groups. For example, an association rule might assert that people who buy peanut butter are 80% likely to also buy jelly
Market basket data is usually transactional. In transactional data, a case is a transaction and the data for a transaction is stored in multiple rows. Oracle Data Mining association models can be built on transactional data or on single-record case data. The ODMS_ITEM_ID_COLUMN_NAME
and ODMS_ITEM_VALUE_COLUMN_NAME
settings specify whether the data for association rules is in transactional format.
Note:
Association models are the only type of model that can be built on native transactional data. For all other types of models, Oracle Data Mining requires that the data be presented in single-record case format.
The Apriori algorithm assumes that the data is transactional and that it has many missing values. Apriori interprets all missing values as sparse data, and it has its own native mechanisms for handling sparse data.
See Also:
Oracle Database PL/SQL Packages and Types Reference for information on the ODMS_ITEM_ID_COLUMN_NAME
and ODMS_ITEM_VALUE_COLUMN_NAME
settings