Hello Philippe,

First of all, thank you very much for sharing the implementations of the algorithms!

My question is the following:

Is there somewhere in the library implementations with rule interestingness measures (e.g., lift)? (I would be interested in sequential rules).

Thank you in advance,
Munira

Hi Munira,

Thanks. I am happy that the library is useful. The goal of this library is to help other researchers by sharing implementations ;-)

Most algorithms for sequential rule mining in the library use the support and confidence measure to evaluate the interestingness of rules.

However, the CMDEO algorithm will indicate not only the support and confidence but also the lift of rules in the output file.

Note that you could also modify other algorithms to calculate the lift or other measures. I think this would not be so difficult.

Best,

Philippe

Thank you very much for your reply, Philippe!

Indeed, CMDeo contains lift calculations. Somehow I've missed that.

And, yes, I will try to implement other measures myself.

Wish you great inspiration for your library! This is really a wonderful resource for learning.


Munira

Great. If you want to share back some code to SPMF after you add some other measures, it could be great if the code is well implemented!

Best regards,