Mining interesting itemsets from transactional database

Association rule mining is a standard technique used for finding the relationships among the itemsets in a database. The method of extracting the frequent itemsets from the database using existing algorithms has several disadvantages such as generation of large number of candidate itemsets, increase in computational time and database scan. With this aim, the paper proposes Mining Interesting Itemsets (MIIS) algorithm which combines the features of partition algorithm and FP tree which reduces the database scan and produces the reduced itemsets from the transactions. The reduced itemsets are validated using the mathematical measures.