Sequential pattern mining is an interesting data mining problem with many real-world applications. Though, new applications introduce a new form of data called data stream, no study has been reported on mining sequential patterns from quantitative data stream. This paper presents a novel algorithm, for mining quantitative streams. The proposed algorithm can mine exact set of fuzzy sequential patterns in fuzzy sliding window and gap constraints entailing the most recent transactions in a data stream. In addition, the proposed algorithm can also mine non-quantitative or transaction-based sequential patterns over a data stream. Numerical results show the running time and the memory usage of proposed algorithm in the case of quantitative and customer-transaction-based sequence counting are proportional to the size of the fuzzy sliding window and gap constraints.
data stream; fuzzy sequential pattern mining; fuzzy constraint; sliding window.