Practical advises on recommendation systems in production

I think it's strange recommendation systems don't ask more about the context.

The answer to 'How do you feel today' can result in a total different playlist.

Or a python programmer who also likes snakes might want to set the context before a search (in DDG you can).

What does this mean? To me it implies logistic regression is already so good that its hard to see the difference between it and most recommender systems. In which case, unless you're operating at massive scale, you can just go with logistic regression

That at least lets the user run the experiment on themselves to check if the recommendation algo makes them sad.

The fact that the sentence above is based on a real thing Facebook did to its unwitting users speaks volumes about our shitty timeline.

Advise is a verb. It reads as if an organization called "Practical" is advising on something.

Edit: advice, no need for plural as in advices

This is indeed excellent new.