A Novel Algorithm for Discovering Frequent Closures and Generators

Abstract

The Important construction of many association rules needs the calculation of Frequent Closed Item Sets and Frequent Generator Item Sets (FCIS/FGIS). However, these two odd jobs are joined very rarely. Most of the existing methods apply level wise Breadth-First search. Though the Depth-First search depends on different characteristics of data, it is often better than others. Hence, in this paper it is named as FCFG algorithm that combines the Frequent closed item sets and frequent generators. This proposed algorithm (FCFG) extracts frequent itemsets (FIs) in a Depth-First search method. Then this algorithm extracts FCIS and FGIS from FIs by a level wise approach. Then it associates the generators to their closures. In FCFG algorithm, a generic technique is extended from an arbitrary FI-miner algorithm in order to support the generation of minimal non-redundant association rules. Experimental results indicate that FCFG algorithm performs better when compared with other level wise methods in most of the cases.