Extended Two-Phase Algorithm for Fast Discovery of High Utility Itemsets on Probabilistic Databases and Possible Application of Counting Inference Approach

In this paper, presented an extension to Two-Phase algorithm that is extended to mine probabilistic high utility patterns, to efficiently prune down the number of candidates and precisely obtain the complete set of high utility itemsets, transaction weighted utility, high utility, counting inference, and probability threshold are used. It performs very efficiently in terms of speed and memory cost both on synthetic and real databases, even on large databases that are difficult for existing algorithms to handle when the algorithm is run using counting inference when it is applicable. Association rule mining algorithm is enhanced for business application.