Mining Weighted Association Rules in Data Streams with a Sliding Window

Association rule mining is one of the most important data mining techniques.Typical association rules consider each item in transactions with the same significance.In order to represent significances of items, every item has be assigned with a weight, and the mining weighted association rules have been proposed.All of the literature on weighted association rules mining, to our best knowledge, is confined to the traditional, relatively static database environment; no research work has been conducted on mining weighted association rules over data streams.In this paper, we propose an algorithm for mining weighted association rules over data streams.Experiments on the synthetic data stream are made to show the effectiveness and efficiency of the proposed approach.