Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

PCRL handles unknown users and/or items #451

Open
wants to merge 4 commits into
base: master
Choose a base branch
from

Conversation

vlainic
Copy link

@vlainic vlainic commented Nov 22, 2021

Description

I have changed recom_pcrl.py following the example of recom_mcf.py, as they have the same modality. I run a test locally on my data and it worked fine.

I did this following #444 as I could not understand why PCRL does not handle this, while all others I have tried to. You made all models just raise ScoreException and then make it a default score. After reading the PCRL paper, I could not understand why it couldn't handle unknown users or items as any other model here. TBH, I think that all models with modalities should somehow handle cold-start, even with the global mean.

Related Issues

As I mentioned above, this is related to the Issue #444 .

Checklist:

  • [-] I have added tests.
    • I think there is no need to add a test.
  • [-] I have updated the documentation accordingly.
    • Same here, as handling unknown parameters is not documented anywhere :)
  • [/] I have updated README.md (if you are adding a new model).
  • [/] I have updated examples/README.md (if you are adding a new example).
  • [/] I have updated datasets/README.md (if you are adding a new dataset).

I just used the same code construction as for MCF.score, as they both have item graph modality. The local tests worked.
@tqtg tqtg requested a review from saghiles November 24, 2021 09:43
I had some bugs too on my side when deploying, and this fixed the problem. Hopefully this review will not fail.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant