earticle

영어

To considered the problem of discovering of interesting association rule among item sets in data base. Algorithms for mining association rule are practical methods to find interesting rules implied in large database. Proposed an innovative approach, beyond minimum support and minimum confidence framework, some extra measures consider for rule improvement is user interestingness constraint. It use three user defined constraint minimum support, minimum confidence, and interesting item and in addition makes use of k- nonempty subset generation methodology of the item which are user interest. Proposed algorithm fundamentally different from the identified algorithms, a number of algorithm is developed for association rule mining, the identified algorithms go through as of number of scanning of data base, and generate the candidate item set , unnecessary or uninteresting rule . The current method applies the user interesting constraint to generate only interesting association rule in data base. Proposed approach not just reduces the number of scanning of data base but also generated frequent itemset, and mine interesting association rule. Experimental result shows that the number of uninteresting rules can be reduced successfully and the validity of rules which mined are better.