Context-Aware RecSys

[tripledoubleE]

I want to develop a recommendation engine using the RecBole library in Python. The characteristic of my data is as follows:

• I have customers who purchase the hotels, so the customers are the users and the hotels are the items to be recommended. Users and hotels have their own features. I want to use these features in my model. I don't have any ratings or reviews of users. That is why, I will recommend hotels to users implicitly.
• Of course, the model should use the interaction between the user and the item as well as their features.
• I tried to use context-aware models, which request .inter, .user and, .item atomic files. But as I don't have any label (rating or review) I couldn't run it properly. (I got KeyError: 'label')
  Can you recommend me, models that I can use in the RecBole library, which uses user and item features and the interaction between user and item, please?

[zhengbw0324]

We suggest that you perform negative sampling on the data before using the context-aware model.

[tripledoubleE]

thanks a lot