Abstract
Sequential pattern mining (SPM) is an important data mining task of discovering time-related behaviours in sequence databases. Sequential pattern mining technology has been applied in many domains, like web-log analysis, the analyses of customer purchase behaviour, process analysis of scientific experiments, medical record analysis, etc. Increased application of sequential pattern mining requires a perfect understanding of the problem and a clear identification of the advantages and disadvantages of existing algorithms. SPM algorithms are broadly categorized into two basic approaches: Apriori based and Pattern growth. Most of the sequential pattern mining methods follow the Apriori based methods, which leads to too many scanning of database and very large amount of candidate sequences generation and testing, which decrease the performance of the algorithms. Pattern growth based methods solve all above problems and in addition, it works on projected database which minimize the search space. Paper reviews the existing SPM techniques, compares various SPM techniques theoretically and practically. It highlights performance evaluation of each of the techniques. Paper also highlights limitation of conventional objective measures and focused on interestingness measures. Finally, a discussion of the current research challenges and pointed out future research direction in the field of SPM.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call