Survey of the study on frequent pattern mining in data streams

Abstract

Data mining and knowledge discovery in data streams have recently attracted more attentions for their applications to numerous types of data, including Web clickstreams, sensor networks, etc. Because of some special characteristics, such as continuous arrival in multiple, rapid, time-varying, possibly unpredictable and unbounded, data streams have yielded some fundamentally new research problems. Among the various topics in this research field, it is paramount to find frequent patterns in data streams in a single pass, or a small number of passes, while using less space of memory. This survey reviewed the last advances in the study on frequent pattern mining in data streams, especially classified the present mining algorithms for the first time and discussed them in detail, and finally suggested some promising research directions in the future.