On a mathematical relationship between the fixed point and the closed itemset in association rule mining

Da Ruan, Tian-rui Li, Tianmin Huang, Yang Xu, Dirk Maes

    Research outputpeer-review

    Abstract

    Association rule mining is one of the important research topics in knowledge discovery and data mining. Recent promising direction of association rule mining is mainly to mine closed itemsets. Based on the Galois closed operators, a mathematical relationship between the fixed point and the closed itemset in association rule mining is discussed and several properties are obtained. To mine all frequent closed itemsets is equal to build the fixed point lattice and mine its all points that satisfy support constraints. A new method for visualization of association rules based on the generalized association rule base is also proposed.
    Original languageEnglish
    Title of host publicationAI 2005: Advances in Artificial Intelligence
    Place of PublicationSydney, Australia
    Pages989-992
    StatePublished - Dec 2005
    EventAUS-Ai 2005 - 18th Australian Joint Conference on Artificial Intelligence - Sydney
    Duration: 5 Dec 20059 Dec 2005

    Conference

    ConferenceAUS-Ai 2005 - 18th Australian Joint Conference on Artificial Intelligence
    Country/TerritoryAustralia
    CitySydney
    Period2005-12-052005-12-09

    Cite this