Examples regarding implicit interactions

[lpoug]

Hello!

First of all thanks for the library, seems like the right successor to tensorflow recommenders and such..

I noticed that all the examples are regarding explicit ratings interactions. Is there any plan to add examples for implicit interactions in the future?

Thanks in advance for the reply.

[abheesht17]

Great suggestion, will work on it! Are you looking for any specific examples in particular? Can tailor the examples on keras.io accordingly.

[lpoug]

Thanks for the reply! Since you asked I do have a suggestion (totally biased toward my own needs of course!). I think an example along the line of an e-commerce recommender system (maybe using the H&M or Criteo datasets). I think this would be highly valuable with one or two examples such as:

• A basic retrieval model like in the Recommending movies: retrieval
  that maximizes Recall@K
• A ranking model (bonus: that can includes additional features like user/item attributes or context features) similar to the Ranking with Deep Learning Recommendation Model
  example, that maximizes Precision@K or any other metric of choice

Apart from that I think the example should illustrate how negative sampling is applied for these types of models based on implicit data (purchases, interactions, ...)