Designing Algorithm on Mining Frequent Patterns in Data Streams
Ao Fu · 2008
This paper introduces some concepts and definitions about frequent patterns in data streams,and presents the general data streams processing model for mining frequent patterns in data streams,and detailedly sums up three classifications of the algorithms on mining frequent patterns in data streams,including Window Model,the Type of Result Set and the Accuracy of Result Set. Based on these classifications,the cube for designing algorithm on mining frequent patterns in data streams is presented. The cube not only covers the exiting algorithms,but also allows for the design of new ones suited for various application requirements. Based on the cube,we analyse some valid strategies for designing the algorithm,aiming at presenting a powerful reference for designing new ones. Lastly,we discuss some future research works about mining frequent patterns in data streams.