An Efficient Algorithm for Discovering Maximum Length Frequent Itemsets

The exploitation of frequent itemsets has been restricted by the the large number of generated frequent itemsets and the high computational cost in real world applications. Meanwhile, maximum length frequent itemsets can be efficiently discovered on very large datasets and are useful in many application domains. At present, LFIMiner_ALL is the fastest algorithm for mining maximum length frequent itemsets. Exploiting the optimization techniques in LFIMiner_ALL algorithm, we develop the MaxLFI algorithm to discover maximum length frequent itemsets by adding our conditional pattern base pre-pruning strategy and evaluating initial length of maximum length frequent itemsets to prune the search space. Experimental results on real-world datasets show that our proposed algorithm is faster than LFIMiner_ALL algorithm for mining maximum length frequent itemsets.