Dear Professor,
I hope you are doing well

I would like to know how to convert a real dataset to be used in FIM.

Is there a standard way that you recommend for me?

Thanks so much for a such nice website.

Best Deniz

Dear Deniz,

Thanks for your message.

There is not really a standard way to convert the data. The reason is that there are a wide variety of data depending on the applications. So how to encode your data in a way that is meaningful for your application depends on what you want to do.

If your data is a text for example, you may encode each sentence as a sequence of items (where items are words) to obtain a sequence database and apply some sequential pattern mining algorithms.

Or if your data is about sequence of actions in an e-learning environments, then you could transform this into a sequence of actions where items represent actions.

There are really many possibilities depending on your data.

This is why in SPMF the format is just a text file and items are represented by numbers. This allow the software to be used in many applications. The only task though is that the data need to be transformed to that format.

Hope this helps

Best regards,

Philippe