search

giuliowaitforitdavide / recsyslearn

GNU General Public License v3.0
7 stars 0 forks source link

Segmenting a dataset: info in the README #3

Closed mmosc closed 2 years ago

mmosc commented 2 years ago

Hi @giuliowaitforitdavide ! In the README, you write

If you don't have the info about the groups, you can use the library itself to segment the dataset. The dataset has to be in the form of user, item, rank.

but I think it should be

If you don't have the info about the groups, you can use the library itself to segment the dataset. The dataset has to be in the form of user, item, timestamp.

Is that right? Cheers, Marta

giuliowaitforitdavide commented 2 years ago

Actually I think that both could be written, because in the case of explicit feedbacks ('user', 'item', 'rank') we use the dataset as it is, but in the case of implicit ones ('user', 'item', 'timestamp') we consider repeated interactions as one interaction. Do you agree? I can fix it in the already created fix/docs branch and I will also have to change the code

mmosc commented 2 years ago

Aaaah, now I see what you meant! I think this is kind of confusing, since 'rank' is used for the ranked lists...

What about using the word 'rating' for explicit feedback?

giuliowaitforitdavide commented 2 years ago

I like it, I will open a branch later

giuliowaitforitdavide commented 2 years ago

I fixed it in the fix/docs branch. I'm opening right now a pull request which I would like you to evaluate. You will see what I wrote in the usage.rst file inside the docs folder