3 ms·
Rec systems that tech giants use looks vastly different than what ML academics study. Industry rec systems are a constantly evolving beast. Why? Because these t
by eachro 5y ago
Rec systems that tech giants use looks vastly different than what ML academics study. Industry rec systems are a constantly evolving beast. Why? Because these teams of ML engineers are constantly running A/B tests to tweak some aspect of the model. The resulting model likely doesnt fit nicely into something category of model you can describe as collaborative filtering/lambdamart/etc b/c it's some highly performant glued together mess.
- w1nk 5y agoYour point about academia studying different systems than industry rings quite true. Lots of the academic recommender systems are built on datasets and techniques that simply don't scale to real world data/applications. That said, this pytorch work looks to live more in the realm of application than academia. Building and scaling large neural embeddings is pretty close to industry practice these days and this library at least claims to solve some of the challenges in doing so.