The task of searching for interesting relationships among data has been always a researching focus in data mining. The overall performance of mining association rules is determined by discovering large itemsets, i.e., sets of itemsets that have their support above a pre-determined minimum support. The different algorithms proposed for association rules task show different approaches to generate all large itemsets: Apriori, AprioriTid, AprioriHybrid, DHP, DIC, Partition, FP-Growth and EquipAsso.
In this paper, the performance of EquipAsso, an algorithm for discovering large itemsets, based on two new operators of relational algebra, is evaluated in relation with Apriori and FP-Growth algorithms, on Tariy, a tool for the Association task loosely coupled with a DBMS.
|