The Facebook Algorithm Mom Problem • Hacker News Board

Re: The Facebook Algorithm Mom Problem

#101

Post by Semaphor » Wed, Jul 12, 2017, 5:47 PM UTC

ryanbrunner wrote: post #95

I find a lot of sites feel like they're overtuning their recommendation engines, to the detriment of using the site. YouTube is particularly bad for this - given the years of history and somewhat regular viewing of the site, I feel like it should have a relatively good idea of what I'm interested in. Instead, the YouTube homepage seems myopically focused on the last 5-10 videos I watched.

I'm not sure if it is because of weird tastes or something else. But for sites like Steam, YT, Netflix I always wish for more tuning parameters because their recommendations are all horrible.

On Netflix, for example, I get recommended stuff that's similar to other things I didn't like.

Amazon is often the best of them for books because their engine seems to only really value the last few things I looked at / bought and so just recommends really similar stuff. Of course, that fails hard whenever I look at something I'm not interested in for whatever reason.

On Steam I wish they had some elaborate filtering system, there are so many games, I need something like NOT(Adventure+Puzzle) BOOST(RPG), I need to combine tags, not just filter them out by themselves.

Re: The Facebook Algorithm Mom Problem

#102

Post by compiler-guy » Wed, Jul 12, 2017, 5:50 PM UTC

jayvanguard wrote: post #78

mrleinad wrote: post #14

I'd like Facebook to have an option to see all posts, without filtering, just as they're posted. It's not hard, it's a simple UX, but it's just not there.

It's probably a core part of their scalability that they never have to produce a complete list of posts. Every fetch is on best effort basis within a small time constraint.

They don't even produce complete search results. Go to a group with a long history, and use "search this group" for a keyword that will return over, say, 100 results--some of them a couple of years old.

You won't see many of the older posts, and facebook gives no indication that they have been pruned. It's very hard to find old things without just scrolling through chronologically, which is slow and error prone.

Drives me insane.

Re: The Facebook Algorithm Mom Problem

#103

Post by ThrustVectoring » Wed, Jul 12, 2017, 5:51 PM UTC

ryanbrunner wrote: post #95

I find a lot of sites feel like they're overtuning their recommendation engines, to the detriment of using the site. YouTube is particularly bad for this - given the years of history and somewhat regular viewing of the site, I feel like it should have a relatively good idea of what I'm interested in. Instead, the YouTube homepage seems myopically focused on the last 5-10 videos I watched.

Youtube is definitely over-weighing the last few videos. Then again, it's likely optimized for a different market segment. Especially with how many children are given tablets. Like, brief intense fascination with many subjects in general is genuinely hard to optimize for, especially when a huge portion of your market has long-lasting intense fascination with a few subjects.

Re: The Facebook Algorithm Mom Problem

#104

Post by mullingitover » Wed, Jul 12, 2017, 5:51 PM UTC

I generally hold facebook in contempt for the forced filtering that they subject me to. Making the 'sort posts chronologically' flag come unstuck is a dirty hack that they should be ashamed of.

Re: The Facebook Algorithm Mom Problem

#105

Post by compiler-guy » Wed, Jul 12, 2017, 5:53 PM UTC

For every social media site XX:

Question: "Why does XX do things this way?"

Answer: "Because it increases engagement."

Why does the Facebook algorithm do this? Because it increases clicks. Why does Youtube use autoplay? Because it increases watch time.

For every single social media site.

Re: The Facebook Algorithm Mom Problem

#106

Post by msujaws » Wed, Jul 12, 2017, 5:53 PM UTC

frenchie4111 wrote: post #49

timsayshey wrote: post #29

Earlier quoted context omitted.

Not all heroes wear capes -- thank you!

Someone should make a browser extension that will automatically (or provides a button to quickly) load the cached version when a site is down Edit: I just made a super simple bookmarklet that does it: http://mikelyons.org/2017/06/12/Google-Cache-Javascript-Shor...

https://testpilot.firefox.com/experiments/no-more-404s/

Re: The Facebook Algorithm Mom Problem

#107

Post by rhizome » Wed, Jul 12, 2017, 5:55 PM UTC

ThrustVectoring wrote: post #103

ryanbrunner wrote: post #95

I find a lot of sites feel like they're overtuning their recommendation engines, to the detriment of using the site. YouTube is particularly bad for this - given the years of history and somewhat regular viewing of the site, I feel like it should have a relatively good idea of what I'm interested in. Instead, the YouTube homepage seems myopically focused on the last 5-10 videos I watched.

Youtube is definitely over-weighing the last few videos. Then again, it's likely optimized for a different market segment. Especially with how many children are given tablets. Like, brief intense fascination with many subjects in general is genuinely hard to optimize for, especially when a huge portion of your market has long-lasting intense fascination with a few subjects.

YouTube would rather you fall into a trance of continuing your recent viewing habits rather than providing a personalized library. It just makes better business sense.

Re: The Facebook Algorithm Mom Problem

#108

Post by tbcktu » Wed, Jul 12, 2017, 5:55 PM UTC

ryanbrunner wrote: post #95

I find a lot of sites feel like they're overtuning their recommendation engines, to the detriment of using the site. YouTube is particularly bad for this - given the years of history and somewhat regular viewing of the site, I feel like it should have a relatively good idea of what I'm interested in. Instead, the YouTube homepage seems myopically focused on the last 5-10 videos I watched.

I hate this. Several years of nothing but a single genre of music in my history but watch one video about something different and get nothing but recommendations based on that. I have to clean up my watch history daily.

Re: The Facebook Algorithm Mom Problem

#109

Post by chongli » Wed, Jul 12, 2017, 5:55 PM UTC

ryanbrunner wrote: post #95

I find a lot of sites feel like they're overtuning their recommendation engines, to the detriment of using the site. YouTube is particularly bad for this - given the years of history and somewhat regular viewing of the site, I feel like it should have a relatively good idea of what I'm interested in. Instead, the YouTube homepage seems myopically focused on the last 5-10 videos I watched.

If I had to guess, I would say YouTube is attempting to get you to watch new types of videos you've never watched before. By focusing on recent videos you've watched, they can try to convert you from a 1-off to a regular viewer of a particular genre.

Re: The Facebook Algorithm Mom Problem

#110

Post by gervase » Wed, Jul 12, 2017, 5:57 PM UTC

ryanbrunner wrote: post #95

I find a lot of sites feel like they're overtuning their recommendation engines, to the detriment of using the site. YouTube is particularly bad for this - given the years of history and somewhat regular viewing of the site, I feel like it should have a relatively good idea of what I'm interested in. Instead, the YouTube homepage seems myopically focused on the last 5-10 videos I watched.

My supposition is that this is a response to changing tastes over time. However, I think 10 videos or less it probably an overreaction to this problem.

For example, let's say they have a user who watches 100 videos a week, for ease of math. 50 of those videos are in "core" areas of interest - these do not change over time. An additional 35 are in areas of secondary interest: topics which have piqued the viewer's curiosity, but not deeply interested them. We can expect these topics to change every [1,4] weeks. The remaining 15 are referrals or clickbait from other websites.

How can YouTube differentiate between these three classes of videos? The first class will be heavily represented in their subscriptions. Presumably, the viewer will prefer their recommendations to ignore the third class (clickbait). The second class is the hardest, as the user may want these videos surfaced, but then want them to decay over time as their interests change.

I think this is the problem that they are attempting to solve, with varying degrees of success.