Mining Weighted Frequent Itemsets Using Window Sliding over Data Streams

In this paper, we considers the problem of mining with weighted support over a data stream sliding window using limited memory space. The continuous characteristic of streaming data necessitates the use of algorithms that require only one scan over the stream for knowledge discovery. This paper focuses on research issues concerning mining frequent itemsets in data streams and we suggests an efficient algorithm WSFI-Mine to mine all frequent itemsets. Our experiment show that our algorithm not only achieved effectively consumes less memory, but also runs significantly faster than THUI-mine.