We propose a novel method for frequent sequence mining over transaction stream (FSM-TS). Unlike itemset mining, the task of FSM-TS is required to handle sequences of itemsets and thus involved in the highly combinatorial explosion of mining objects. The proposed method addresses this problem by extending two key techniques, called incremental intersection and resource-oriented approximation, in the context of FSM-TS. In this paper, we briefly describe the validness of it and show a preliminary

more »

... esult in the experiment obtained using real datasets.