The Algorithm of Mining Frequent Closed Itemsets Based on Index Array

The set of frequent closed itemsets determines exactly the complete set of all frequent itemsets and is usually much smaller than the latter. In this paper, an algorithm based on index array for mining frequent closed itemsets, Index-FCI is proposed. The vertical BitTable is adopted to compress the dataset for counting fast the support. To make use of the horizontal BitTable, the index array corresponding to the database is constructed and a new concept GFI (Great Frequent Itemset) is defined, which can be quickly found from the index array to reduce the closed itemsets checking. The hash table whose hash function value is the support of the itemset is created to remove any frequent but “non-closed” itemsets using the hash pruning. In Index-FCI, the database is firstly compressed into BitTable; secondly, the index array corresponding to the dataset is constructed; thirdly, GFI is found from the index array and the hash table is created to store frequent closed itemsets; finally, the hash table is traversed to obtain all frequent closed itemsets. Experimental results show that Index-FCI is suitable for mining frequent closed itemsets.