Organizing the discovered association rules based on general-specific (GS) hierarchical patterns

Many existing association rules mining algorithms techniques often produce a large number of rules, which make it very difficult for the user to analyze them manually. The key problem is not with the large number of rules because if there are indeed many rules that exist in data, they should be discovered. The main problem is with our inability to organize and represent the rules in such a way that the user can easily comprehend them. This paper proposed a technique to intuitively organize the discovered association rules in a hierarchical fashion named general-specific (GS) pattern. With this organization, the user can view the association rules at different level of details. The technique first finds a subset of the association rules called the most-general rules set (MGRS) to give the user a general relationship or a big picture of the original rules set. Then, the user can selectively view more-specific rules below a general rule that are interesting to him/her. Experiment results and practical applications show that the technique is both intuitive and effective.