Upselling recommendation with NDR for implicit feedback (millions users, few items)

[lpiscusc]

I am currently developing an upselling recommendation system.

Scenario:

• Millions of users
• Few dozen items
• Most users do not interact with any items
• No ratings, only known whether the user has interacted with the item (implicit feedback)

Approach:

• Implementing an NDR (Two-tower) architecture using user and item features
• Currently treating it as a retrieval problem, inputting only user-item interactions (positive feedback) to the model

I was wondering if this is the right approach or if it might make sense to transform it into a ranking problem (1 if the user interacts and 0 otherwise), with the caveat that the size of the database could explode because if I understand correctly, negative examples would have to be created manually.
What is the best way to proceed?

[rlcauvin]

Would you provide more information about how items are presented to users? Do you have data on the implicit negative interactions, events where a user had an opportunity to engage positively with an item but chose not to do so?