On: Categories of deep recommendation systems

[timkam]

There are recent works that question whether "deep" learning algorithms, and in particular neural networks, are really advancing the state of the art of recommender systems: https://arxiv.org/pdf/1907.06902.pdf (Best Paper Award, RecSys 2019). It is strange that the blog post fails to mention this.

[amznthrowaway5 → timkam]

Yeah, this has been a well known problem in the field for a while now, and not just with DNN approaches https://en.wikipedia.org/wiki/Recommender_system#Reproducibi...

These systems almost never work in production. The publications are usually very deceptive.