MAXFP: A Multi-strategy Algorithm for Mining Maximum Frequent Patterns and Their Support Counts
Trends in Applied Sciences Research,
2006, 1(4), 402-415.
The problem of efficiency is the main crux of most data mining problems, such as mining frequent patterns. This problem is mainly concerned with number of operations required for counting pattern supports in the large database. In this study we propose a Multi-strategy based new algorithm, which combines Pincer-search and counting Inference approaches for discovering maximum frequent patterns along with support count of all their subsets. This algorithm works in both directions, bottom-up as well as top-down. The main search direction is still bottom-up but a restricted search is conducted in the top-down direction. The important characteristic of the algorithm is that, it is not necessary to explicitly examine every frequent itemset. Counting inference allows us to perform as few support counts as possible. Using this method, the support of a pattern is determined without accessing the database whenever possible using the supports of some of its sub-patterns called key patterns. MAXFP method performs well even when some maximal frequent itemsets are long. It reduces cost of the frequent itemsets discovery process, that is minimizes support count operations as well as database scans and it count support of all frequent patterns which are generated by maximum frequent patterns, without accessing database.
Cited References Fulltext