Abstract: Real-world applications often generate huge amount of data streams, which challenges efficient processing and mining due to its special characteristics. As a fundamental problem in data stream mining, frequent pattern mining techniques employed in these applications should be efficient in terms of space usage and execution time while providing a high quality of yields. This has received considerable attention in the past few years due to its research value and increasing amount of importance in numerous applications. The purpose of this paper is to review the recent work in frequent pattern mining under data stream environments, and summarize its characteristics and algorithms in general. With taxonomy, we dissertate the existing algorithms from probabilistic and deterministic bounds on error, respectively. Throughout the detailed review, some comparisons and evaluations are performed. Finally, future directions in data stream mining research are discussed.