A New Closed Frequent Itemsets Mining Algorithm Based on GPU

摘要

Vertical data format is an important structure in closed frequent itemsets mining. All closed frequent itemsets can be found by simply using the operations of "and" and "or". But it consumes huge amount of storage space, especially in the case of large size data set. This paper proposed an algorithm for mining closed frequent itemsets based on a new data structure deriving from the vertical format. This new data structure can contribute to save storage space by using a multi-layer index. When dealing with large data sets with the acceleration of GPU, this algorithm can obtain a high speed. Our experimental results show that our algorithm uses much less computation time than other similar methods.