I want to find top-k frequent sequential patterns from a list of sequences. Order of occurrence matters here (subsequence (1,2) is not same as (2,1)). Suppose I have 2 sequences:

  • S1=[1,2,3,1,5,2]
  • S2=[3,4,1,5,3,2]

If I use sequential pattern algorithm like prefixspan then sup(1,2)=2 (it occurs in both S1 and S2). But I want an algorithm to calculate all occurrence including repetition in the same sequence i.e. sup(1,2)=3 (2 times in S1 and 1 time in S2) and gap is allowed here (in [...1,5,2...] the subsquence (1,2) is counted 1).

Is there an sequential pattern mining algorithm that do exactly that Or is it no longer sequential pattern mining?

If it is not then what should I use to solve this problem? I am planing to use python to solve this problem so if you know any library that can do exactly this please also give me some recommendations. I am currently using spmf but as I explained sequential pattern algorithm don't exactly match what I need.

Thank you for your help!



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.