Efficient reduction of the number of associations rules using fuzzy clustering on the data

摘要: In this paper, we are interested in the knowledge discovery methods. The major inconveniences of these methods are: i) generation a big number association rules that not easily assimilated by human brain ii) space memory and time execution necessary for management their data structures. To cure problem, propose to build (meta-rules) between groups (or clusters) resulting from preliminary fuzzy clustering on data. We prove can deduce about initial set if want more details. This solution reduced considerably generated rules, offered better interpretation optimized both time. approach is extensible; user able choose or extraction algorithm according domain his needs.