An Efficient Frequent Closed Itemsets Mining Algorithm Over Data Streams

Mining frequent closed itemsets provides complete and condensed information for frequent pattern mining. Online mining of frequent closed itemsets over streaming data is one of the most important issues in mining data streams. In this paper, we first present a general methodology to identify closed itemsets over data streams, using concept lattice theory. Using this methodology, we then proposed a novel sliding-window based algorithm which is based on concept lattice incremental restructuring and lattice construction. The algorithm exploits lattice properties to limit the search to frequent close itemsets which share at least one item with the new transaction. A thorough performance study on synthetic datasets has shown that our proposed algorithm is both time and space efficient and adapts very rapidly to the change in data streams.