Transcript DATA MINING
DATA MINING Using Association Rules by Andrew Williamson What is Data Mining? • A.K.A. – Knowledge Discovery in DataBases (KDD) • “Data mining is the automated extraction of hidden predictive information from databases. Data mining software allows users to analyze large databases. This can enable you to solve business decision problems, for data mining as an extension of statistics; while it doesn't solve your problems, it is the technology that can find your problems. You can build a predictive model of an ideal customer from your own databases and, using this information, build your marketing strategy or even determine the creation of new products and services.” • - www.biz-services-savvy.com • In short “Data mining can be defined as, “the non-trivial extraction of implicit, previously unknown, and potentially useful information from data” ”. An example of Association “A simple example of data mining is its use in a retail sales department. If a store tracks the purchases of a customer and notices that a customer buys a lot of silk shirts, the data mining system will make a correlation between that customer and silk shirts… Association example cont. …The sales department will look at that information and may begin direct mail marketing of silk shirts to that customer, or it may alternatively attempt to get the customer to buy a wider range of products. In this case, the data mining system used by the retail store discovered new information about the customer that was previously unknown to the company. “ • - wikipedia.org Association Rules • Using Association Rules, X Y, helps to identify shopping trends over a given amount of time • 1) Support • 2) Confidence Association Rules Cont. • 1) Support – The ratio of transactions that contain both an item X and an item Y, over all transactions. • P(X U Y) – Measures the significance of the given rule. Association Rules.. • 2) Confidence – The ratio of transactions containing X, and also contains Y. • P(X U Y) / P( X ) – Measures the strength of the correlation of the given rule. Association Rules… • The Association rules X Y with a support and confidence ratio of 50% or more are considered to be meaningful and therefore are kept, otherwise the rule will be discarded or ignored. 2 Phases of Association Rules • When considering an association rule to examine follow these 2 steps: • Phase 1 – List together all rules with a high support ratio. – A.K.A. – The Frequent Itemset. • Phase 2 – List together all rules with a high Confidence ratio from the Frequent Itemset group. Extended Association Rules • Standard Association Rules express a correlation between values of a single dimensional schema. • While, more meaningful associations may be discovered when incorporating multi dimensional schemas Extended Association Rules cont. • A Simple Example: – Sleeping bags Tents • This alone, seems obvious and not useful. – Sleeping bags Tents ( region=north, season=summer) • more meaningful incorporating region and season of transactions. Extended Association Rules cont.. • More meaningful information may be associated by augmenting the original Association Rules – X Y (Z) • Transactions which satisfy Z and contain X, are more likely to contain Y as well. Work Cited • • • • www.wikipedia.org http://www.biz-services-savvy.com http://sba.luc.edu http://www-128.ibm.com