FastLMFI: An Efficient Approach for Local Maximal Patterns Propagation and Maximal Patterns Superset Checking

TL;DR

FastLMFI introduces an efficient indexing method for local maximal pattern propagation and superset checking, significantly improving performance over existing algorithms in mining maximal frequent itemsets, especially on dense datasets.

Contribution

The paper presents FastLMFI, a novel indexing approach that enhances maximal pattern propagation and superset checking, outperforming prior techniques and integrating with state-of-the-art algorithms.

Findings

Outperforms previous methods on dense datasets.

Improves maximal itemset mining efficiency.

Effective in pruning search space.

Abstract

Maximal frequent patterns superset checking plays an important role in the efficient mining of complete Maximal Frequent Itemsets (MFI) and maximal search space pruning. In this paper we present a new indexing approach, FastLMFI for local maximal frequent patterns (itemset) propagation and maximal patterns superset checking. Experimental results on different sparse and dense datasets show that our work is better than the previous well known progressive focusing technique. We have also integrated our superset checking approach with an existing state of the art maximal itemsets algorithm Mafia, and compare our results with current best maximal itemsets algorithms afopt-max and FP (zhu)-max. Our results outperform afopt-max and FP (zhu)-max on dense (chess and mushroom) datasets on almost all support thresholds, which shows the effectiveness of our approach.