List of keywords in as tabular data?

fastai
1

Hi!

How can I handle lists in tabular data? I am trying to implement this example https://docs.fast.ai/tabular.html on my own dataset.

I have one column called keyword that contains a list of keywords. How should I handle this in a tabular data sense? Should i convert it to keyword_0, keyword_1, keywords_2… and have some ‘Nan’?
it is not a consistent number of keywords either, it varies from 3-10.

any pointers is much appreciated, thanks! :slight_smile:

2

It depends on how you want them handled: if you leave them as lists, you will have one category per possible value of your list. It might make more sense to separate them in different columns, yes.

1 Like