Association models are often used to perform market basket analysis to discover relationships or correlations among a set of items. Such models are widely used in data analysis for direct marketing, catalog design, and other business decision-making processes. A typical association rule of this kind asserts the likelihood that certain behaviors are associated, for example,"70% of the people who buy spaghetti, wine, and sauce also buy garlic bread."
Association models are designed to process sparse data; indeed, if the data is not sparse, the algorithm may require a large amount of temporary space and may not be able to build a model. Market basket data is always sparse, that is, a small fraction (always less than 20%, possibly as little as 3%) of the attributes are non-zero or non-null in a given case. For example, in a small grocery store, there might be 10,000 products in the store with the average size of a basket (the collection of distinct items that a customer purchases in a typical transaction) 50 or fewer items.
Note: Oracle Data Miner requires that the input data for an Association model be in transactional format. This is a restriction of the formats supported by the Oracle Data Mining programmatic interfaces. For information about table formats, see Table Formats.
Oracle Data Mining calculates the following two properties for the rules generated by an association model:
Oracle Data Mining calculates associations using the Apriori Algorithm.
Copyright © 2006, 2008, Oracle. All rights reserved.