Closed rudf0rd closed 7 years ago
Yes at present the closest would be to have separate meta-data fields actor1..actorN. For genres probably a boolean field for each.
Most of the current algorithms for recommendation built in are for collaborative filtering where the activity is the core data used to create a Model. For algorithms where you want to add extra context meta-data along side meta data I would look at Factorization Machines https://www.slideshare.net/hongliangjie1/libfm
We don't presently have a FM based model unfortunately.
Reading through the docs, I don't see a way in the meta data structuring to add an attribute with an array data type. Here's my issue:
I want to add ~10 actors per movie and ~15 actors per tv show. I originally thought about denormalizing this into
actor_1
,actor_2
, etc. But that won't work relating separate attributes unless an actor happens to be inactor_1
multiple times, right?In a perfect world, I'd be able to something like:
Is there anything I could do to get close to something like this? Is there something else I'm missing that would work instead?
The same thing happens with genres since a movie or tv show usually has multiple genres (ex: thriller, action vs. thriller, horror).