<![CDATA[As one application of data mining, market basket analysis is generally performed usingApriori method. This method of looking for associations between items by simply countinghow many times the items appear in the overall transaction, regardless of quantity of items inthe transaction. Therefore, the researchers propose a new method based on the perception thatthe more items purchased in a transaction, then the relationship between items in atransaction is getting weaker.To test the realization of the concept, we collect the sales transaction data from a supermarketfor a month. The collected data were converted to a form that can be processed by theprogram. Then, an algorithm created to process data in order to generate association rulesfrom those items in the transaction. Rule can be displayed in the form of tables or graphs.Experiment was made using a software program Borland Delphi and MS Access.By using the output from this program, which is an association rule, we can know the level ofassociations between items that are useful for helping the decision makers in determining themarketing policy. From the test results, we can conclude that the smaller the minimumsupport and confidence are determined, the more rules that can be generated and the time ittakes more and more. In addition, the higher the number of combinations to be searched, theless time is needed.]]>
Copyrights © 2005
