HCG: A new algorithm for mining share-frequent patterns

摘要

The paper proposes a new efficient algorithm, named HCG algorithm, to mine share-frequent patterns from an incremental pattern set table knowledge called PSTable. The PSTable stores all non redundant patterns with their count information by a single database scan. A transaction newly added to the database can be incrementally added to the PSTable. The new algorithm efficiently discovers all share-frequent patterns from the PSTable by generating candidates from only high share atomic patterns. Its correctness is assured by the downward closure property. The experiment results on dense and sparse datasets show that the proposed algorithm is more efficient than existing algorithms in terms of both execution time and number of candidates.