PROJECT TITLE :

Mining High Utility Patterns in One Phase without Generating Candidates

ABSTRACT:

Utility mining could be a new development of data mining technology. Among utility mining problems, utility mining with the itemset share framework could be a laborious one as no anti-monotonicity property holds with the interestingness measure. Prior works on this problem all employ a 2-section, candidate generation approach with one exception that is but inefficient and not scalable with large databases. The 2-phase approach suffers from scalability issue due to the huge number of candidates. This paper proposes a novel algorithm that finds high utility patterns in a very single section while not generating candidates. The novelties lie in a very high utility pattern growth approach, a lookahead strategy, and a linear information structure. Concretely, our pattern growth approach is to go looking a reverse set enumeration tree and to prune search house by utility upper bounding. We tend to also look ahead to spot high utility patterns while not enumeration by a closure property and a singleton property. Our linear data structure enables us to compute a decent bound for powerful pruning and to directly establish high utility patterns in an economical and scalable method, which targets the basis cause with prior algorithms. Extensive experiments on sparse and dense, synthetic and globe data counsel that our algorithm is up to one to three orders of magnitude a lot of economical and is additional scalable than the state-of-the-art algorithms.