Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How useful was the Netflix Prize challenge for Netflix?

quora.com

Post reply on HN
Print view View original

101–110 of 205 posts

Re: How useful was the Netflix Prize challenge for Netflix?

#101

Post by laurent92 »

koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

I suspect their algorithm is stuck in a local minimum, where it has proven to itself that, if a movie is presented to you hundreds of time, there will be a moment where you click, either inadvertently or because there is nothing else presented to you, and it counts as a validation of their engine. It is so optimized for this that it doesn’t try just presenting all movies anymore - which is a recurrent problem in A/B testing in general.

Yes, Netflix’ engine is the reason I left Netflix…

Re: How useful was the Netflix Prize challenge for Netflix?

#102

Post by j7ake »

koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

The great thing with these transparent parameters is that it can be user-adjusted to people's preferences. For example, maybe don't recommend if watched < x months ago.

Re: How useful was the Netflix Prize challenge for Netflix?

#103

Post by jerzyt »

koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

What Netflix is doing is hiding the fact that they have very little new content. If they implemented all your suggestions, I'd have about 3 titles to watch, which I would like, but they're worried that I'd cancel the subscription.

Re: How useful was the Netflix Prize challenge for Netflix?

#104

Post by LaundroMat »

stepbeek wrote: post #100
szszrk wrote: post #98

Earlier quoted context omitted.

In the end they will promote what they WANT to be popular, so our suggested titles will still be a mess of shit we already watched or disliked with a button. Shows with 10 seasons you hated and disliked after 5 minutes will still be in our "continue watching" or suggested.

> In the end they will promote what they WANT to be popular This is a hugely important point. I'm completely uninterested in most netflix originals, but I understand why they're going to continue to recommend them to me.

I'd like a way to filter out the Netflix Originals too.

Re: How useful was the Netflix Prize challenge for Netflix?

#105

Post by amelius »

derangedHorse wrote: post #59
kens wrote: post #51

Earlier quoted context omitted.

> make the Netflix brand exciting and legitimate to engineers As a serious question, why do people include Netflix in the acronym FAANG, which I see on HN all the time? Is there something special about Netflix? Netflix is around the #14 tech company, so it's strange to see Netflix in there instead of Microsoft. Or is the use of FAANG divorced from its literal meaning?

It's probably included because of the sky-high salaries they offer since FAANG is typically an acronym used to refer to top software companies to work for. From what I've heard from friends, Microsoft typically pays the least out of all the companies that make up the acronym and their technologies are also seen as less trendy than the other companies listed.

How can Microsoft be less trendy if they publish way more research?

Re: How useful was the Netflix Prize challenge for Netflix?

#106

Post by TeMPOraL »

koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

I wonder if there's space for a proper, quality recommendation engine for Netflix. In my books, it would do exactly what you described here, plus provide way for user to input persistent constraints (e.g. "Don't show me movies from 'Horror' genre") and queries (e.g. "genre: sci-fi, after: 2010").

How could it work with Netflix without their explicit support? For the movie database, I assume there's some data source somewhere that lists all the movies and series Netflix currently has in any given region. As for ingesting watching history, it could parse Content Interaction History exported from Netflix via their GDPR Subject Access Request flow. Sure, they have up to 30 days to process such request, but I'd happily accept a recommendation system I have to manually update every month, over the disaster Netflix has been offering to its users.

Re: How useful was the Netflix Prize challenge for Netflix?

#107

Post by snissn »

koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

i wish they would expand out their browsing and browsing by genre too!

Re: How useful was the Netflix Prize challenge for Netflix?

#108

Post by manigandham »

koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

There’s not enough content for completely new listings every time, and people do watch things repeatedly, especially TV shows.

Re: How useful was the Netflix Prize challenge for Netflix?

#109

Post by teichmann »

vletal wrote: post #96
koonsolo wrote: post #92

I don't get why they wanted to go all out on AI for this. I use Netflix, and a simple algorithm would already be an improvement: - If I watched a movie and gave it a thumbs down, don't recommend it - If I watched a movie - If I browsed over a movie 50 times, read the info, and still didn't play it, stop recommending it. - If I watched the last episode, remove the "new episodes" banner. WTF

- Do you really get recommenditions for movies you already watched or down voted? I'm not aware of that happening to me. It's not it the case that you use multiple different profiles within a single account? I sometimes see that for shows we finished, yet people tend to rewatch episodes... - Removing a movie based on "not payed X times" would remove all popular movies for all users in a multiple of X steps. -I agree…

I'm not sure about down voted ones, but I can confirm that I sometimes get already watched movies recommended again (even in recommendation mails).

Re: How useful was the Netflix Prize challenge for Netflix?

#110

Post by dooferlad »

stepbeek wrote: post #100
szszrk wrote: post #98

Earlier quoted context omitted.

In the end they will promote what they WANT to be popular, so our suggested titles will still be a mess of shit we already watched or disliked with a button. Shows with 10 seasons you hated and disliked after 5 minutes will still be in our "continue watching" or suggested.

> In the end they will promote what they WANT to be popular This is a hugely important point. I'm completely uninterested in most netflix originals, but I understand why they're going to continue to recommend them to me.

What Netflix wants to be popular also tramples over what I am already watching. I find that the UI gets in the way of enjoying several programs that I am part way through by pushing something new. If they want to continue to make series and only get me to watch half of the episodes then the algorithm is spot on.

That said, Amazon Prime video is so much worse. It suggests that I would like to watch series 1 of a show when I have watched it and am part way through season 2.

Post reply on HN