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

61–70 of 205 posts

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

#61

Post by square_usual »

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.

That's why Microsoft isn't in the acronym!

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

#62

Post by ma2rten »

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.

That is not the origin of the term. It was coined by Jim Cramer for fast growing stocks.

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

#63

Post by dralley »

kuiper0x2 wrote: post #56
ffggvv wrote: post #54

Earlier quoted context omitted.

it was a originally just FANG with one A. and it was just partially coined because it’s a catchy term. it’s also pretty dated at this point. and if you took the N out it would not be appropriate

The N should be Nvidia

You'd need to add another "A" for AMD.

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

#64

Post by rurp »

echelon wrote: post #47
smnrchrds wrote: post #7

Earlier quoted context omitted.

Anything can beat their current algorithm of "push people towards our originals, preferences be damned". EDIT: I should add in terms of customer satisfaction, not revenue. I am sure forcing their originals down people's throats is great for their revenues.

Licensed content costs them more. Users that consume licensed content are worth less if that's all they consume.

Well yeah, that's what is in Netflix's interest, but there's a reason many companies succeeded by focus on improving the user's experience. I guess Netflix has a a different approach.

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

#65

Post by hobofan »

Supermancho wrote: post #25
BB212 wrote: post #20

I think the discussion misses the most important part: The goal of the Netflix prize wasn't to come up with the best algorithm - it was to make the Netflix brand exciting and legitimate to engineers. At the time, Netflix wasn't super high-tech and I'm sure it was hard for them to get the top talent they needed. It seems silly in retrospect now, but I'm certain the reason this was approved was because they wanted the…

> At the time, Netflix wasn't super high-tech I will admit that it was interesting to see what algorithms were poised to be cutting edge in media recommendation. The result was rather disappointing to me. Netflix STILL isn't that exciting from anything but a compensation standpoint. The problems at netflix are about programming, while the technical challenges are droll at best.

Yeah, I'm sure being responsible for 10% of the internet's traffic is trivial on a technical level, together with all the codec and video encoding engineering they are doing.

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

#66

Post by _pastel »

uyt wrote: post #42

> are open algorithmic contests useful and valuable? Kaggle has been around for a long time now. If it works, I would expect them to be pumping out tons of interesting results from winners but I don't think I've heard many stories like that. It seems to be mostly useful for recruiting purposes?

Kaggle competitions rarely produce interesting algorithmic results.

But I highly encourage you to read the winners' solutions. They are full of clever data insight, augmentations, regularizations, feature engineering, and preprocessing and postprocessing tricks.

But above all, compared to the academic literature, it's shocking how much time and creativity they spend on validation. Maybe I'm reading the wrong papers, but the flashy new neural architectures rarely even mention their validation setup; Kaggle winners sometimes devote half of their explanation to it. It's part of their secret sauce.

Two personal favorites:

(1) https://www.kaggle.com/c/severstal-steel-defect-detection/di.... The "random defect blackout" was a really clever data augmentation.

(2) https://www.kaggle.com/c/ieee-fraud-detection/discussion/111.... Particularly how they reduced overfitting with adverserial validation. They trained a separate model to distinguish between train and test sets, and then dropped features that ranked highly in feature importance on that model. That's probably a well-known technique in some circles, but I had never seen anything like it before.

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

#67

Post by rudyfink »

setr wrote: post #28
Supermancho wrote: post #25

Earlier quoted context omitted.

> At the time, Netflix wasn't super high-tech I will admit that it was interesting to see what algorithms were poised to be cutting edge in media recommendation. The result was rather disappointing to me. Netflix STILL isn't that exciting from anything but a compensation standpoint. The problems at netflix are about programming, while the technical challenges are droll at best.

IMO the recommendations are no good because they fundamentally take the wrong approach — rather than ask the user what they like, they try to guess what you like based on usage (which really doesn't correlate well — I watch a lot of garbage because I can’t find things I like, and I don’t have anything better to do.) And they don’t ask because users don’t provide useful answers. But users don’t provide useful answers,…

The recommendation system, historically (i.e., in the long-long ago of spinning disks), was insanely good. But then Netflix moved to streaming and, as a consequence, its own--and generally less good--content.

By analogy, Netflix went from being a sci-fi future of having and being able to recommend on the basis of _everything_, to having a handful of good offerings and a huge amount of b-movie-level offerings.

My gut sense is management tried to paper over this "content loss problem" by making changes:

1) to the recommendation system to push Netflix content[1]; and

2) making changes to the UI to force users to be more reliant on the recommendation system.

I suspect these changes have, generally speaking, made user-consumption metrics look decent--in my mind the core of almost all Netflix's post-streaming decisions. But, as you suggest, it is all papering over a problem of user dissatisfaction: Netflix recommends you mediocre content, and you eventually give up and watch it--and then feel meh.

[1] I can imagine Netflix executives being unwilling to report that the content Netflix had paid mightily for scored low on Netflix's own recommendation algorithm. Philosophically, Netflix went from being, essentially, content agnostic (e.g., it just bought more of X DVD), to having incentives to see particular content (e.g., its own) rank highly.

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

#68

Post by fouc »

setr wrote: post #57
fouc wrote: post #44

Earlier quoted context omitted.

I think you've hit the nail on the head here. There's no good way to express how I feel when I'm watching a show or at the end of the show. There's no "Holy shit, this is amazing" vs "This is decent" etc - where sentiment is clearly attached to the rating. A 5 star or 3 star rating scale alone isn't quite good enough..

I don’t think that’s correct. 1-5 stars is sufficient. The problem is that you need reason to continuously update the values as your preferences update over time (what was once a 5-star is now a 4-star, because that last movie I saw was phenomenal ) What you need is sufficient reason to do so — the values need to actually be useful to you to make updating an act of sanity (unlike now, where it’s purely an act of futi…

Relative rating could also work here, "did you enjoy this movie more than this other movie you recently saw" type of deal.

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

#69

Post by lacker »

kens wrote: post #51
BB212 wrote: post #20

I think the discussion misses the most important part: The goal of the Netflix prize wasn't to come up with the best algorithm - it was to make the Netflix brand exciting and legitimate to engineers. At the time, Netflix wasn't super high-tech and I'm sure it was hard for them to get the top talent they needed. It seems silly in retrospect now, but I'm certain the reason this was approved was because they wanted the…

> 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?

Because FAANG is a cool, somewhat-evil sounding word, and GAFAM is a boring nothing.

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

#70

Post by throwaway3699 »

erklik wrote: post #46
setr wrote: post #28

Earlier quoted context omitted.

IMO the recommendations are no good because they fundamentally take the wrong approach — rather than ask the user what they like, they try to guess what you like based on usage (which really doesn't correlate well — I watch a lot of garbage because I can’t find things I like, and I don’t have anything better to do.) And they don’t ask because users don’t provide useful answers. But users don’t provide useful answers,…

On the other hand, the only platform that provides me with good recommendations to watch things seems to be TikTok. They are not asking me to rate individual videos, and so on. Clearly, there is a way to do recommendations without "ratings".

TikTok has two advantages:

- lots of short content

- viewing metrics to the second

Within an hour of usage you could've browsed through hundreds of TikToks, and allowed them to classify many tastes for you.

You'd need to sit in front of Netflix for an entire month for them to get the same amount of signal.

Post reply on HN