We focus on the problem of mining probabilistic maximal frequent itemsets. In this paper, we define the probabilistic maximal frequent itemset, which provides a better view on how to obtain the pruning strategies. In terms of the concept, a tree-based index PMFIT is constructed to record the probabilistic frequent itemsets. Then, a depth first algorithm PMFIM is proposed to bottom-up generate the results, in which the support and expected support are used to estimate the range of probabilistic support, which can infer the frequency of an itemset with much less runtime and memory usage; in addition, the superset pruning is employed to further reduce the mining cost. Theoretical analysis and experimental studies demonstrate that our proposed algorithm spends less computing time and memory, and significantly outperforms the TODIS-MAX[20] state-of-the-art algorithm.
CITATION STYLE
Li, H., & Zhang, N. (2016). Probabilistic maximal frequent itemset mining over uncertain databases. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9642, pp. 149–163). Springer Verlag. https://doi.org/10.1007/978-3-319-32025-0_10
Mendeley helps you to discover research relevant for your work.