Abstract:

Combat simulation data stream exists a variety of associated between data attributes, so analysis of data stream association rules is designed to find out the hidden relationship. In view of the combat simulation data stream characteristics of large volume data, variable length, this paper studies how to get the frequent itemsets in the the association rules,in order to obtain frequent itemsets from combat simulation data stream within the limited space quickly, in view of space usage shortages of association rules algorithms Lossy Counting , we proposed a kind of the improved frequent itemsets search method (FIBM)based on the matrix, the method use triangular itemsets matrix instead of the trie tree structure in Lossy Couting algorithm, thereby effectively reducing the Lossy Couting algorithm space occupancy rate.Numerical experiments prove that FIBM algorithm for frequent itemsets search are more improved performance obviously than Lossy Couting algorithm.

Key words: equipment combat simulation data stream;associated rules;frequent itemsets search