A improved sequential pattern mining algorithm based on PrefixSpan

In order to solve the problem of large space and time overhead in the PrefixSpan algorithm, a new sequential pattern mining algorithm based on PrefixSpan is proposed, which is PrefixSpan-x algorithm. The algorithm reduces unnecessary storage space and removes the non-frequent items, and combines AC automaton to optimize frequent sub-sequence mining, so as to improve the performance. Experiments prove that the performance of PrefixSpan-x algorithm is better than Prefixspan algorithm in time and space.