ASCF: Optimization of the Apriori Algorithm Using Spark-Based Cuckoo Filter Structure

Table 1

Gaps in the previous approaches addressed by ASCF.

Gaps in the previous approaches	Improvement introduced by ASCF
Working on the MapReduce framework, which causes significant I/O overhead and increases the time cost	Working on the spark framework, which highly improves the performance
Using the Bloom filter to reduce the time of generate 2-frequent itemsets in the second iteration or using a hash tree to store candidate k + 1 itemsets, which takes a very much time and storage space when candidate combinations are too many	Using the cuckoo filter that has only 1-frequent items to prune the transactions and generate candidate sets from those transactions in all the iterations