Performance Evaluation of Iar Algorithm In Discovering Frequent Item Sets

Improving Rule Discovery Performance with IAR Algorithm

Abstract

The generationof Association rule mining is to discover the association rules. The frequentitemsets found in the previous step are used to generate association rules. Allthe permutations and combinations of the items present in the frequent itemsetsare considered as candidates for strong rules. A lot of rules will be generatedin this way. A strong rule is one that has minimum confidence which is computedby the Formula The main difference between Apriori and IAR algorithm is thatIAR algorithm takes user’s attribute preference for the resulting rules.Thereafter, the IAR searches for rules that contain the user specifiedattributes on the L.H.S. and derive other attributes in the database. If such arule possesses high confidence level then it could be valuable in the marketingcontext for the organisation. In this way a lot of time can be saved and theuser trusts more in the discovered rules.