Design of a sequential pattern mining algorithm based on effective items prefixspan
Chen Hu, Youcheng Su, Zilong Wang, Jianyu Wang · 2021
In the algorithm of sequential pattern mining, experts and scholars spent a lot of time and energy to put forward many typical methods, such as: SPADE, GSP (Generalized Sequential Pattern mining algorithm), PrefixSpan algorithm, etc. In this paper, the above algorithm will be analyzed and studied. Through experiments, PrefixSpan is verified to be faster than the other two algorithms under the same conditions. Therefore, PrefixSpan will be used as the research algorithm in this paper. Based on the research on the characteristics of the PrefixSpan algorithm, aiming at the problem of too many projection databases in the process of operation, we optimize the PrefixSpan algorithm, introduce effective item number, put forward EIPrefixSpan algorithm, and optimize the PrefixSpan algorithm in space/time.