DSM-FI: an efficient algorithm for mining frequent itemsets in data streams

摘要: Online mining of data streams is an important problem with broad applications. However, it also a difficult since the streaming possess some inherent characteristics. In this paper, we propose new single-pass algorithm, called DSM-FI (data stream for frequent itemsets), online incremental itemsets over continuous transactions. According to proposed each transaction projected into set sub-transactions, and these sub-transactions are inserted in-memory summary structure, SFI-forest (summary itemset forest) maintaining all embedded in generated so far. Finally, determined from current SFI-forest. Theoretical analysis experimental studies show that algorithm uses stable memory, makes only one pass transactional stream, outperforms existing algorithms one-pass itemsets.