OPAM-An Efficient One Pass Association Mining Technique without Candidate Generation

This paper presents an efficient One Pass Association Mining technique i.e. OPAM, which finds all the frequent itemsets without generating any candidate sets. OPAM is basically an integration of two techniques: a correlogram matrix based technique to generate all the frequent 1and 2-itemset in a single scan over the database and a technique that uses vertical layout concept to generate the rest of the frequent itemsets. The proposed technique also supports interactive mining of 1and 2-element frequent itemsets. We experimented with several synthetic datasets and compared the performance of OPAM with its other counterparts i.e. Apriori and FPgrowth and found satisfactory.