Party in the issues: https://github.com/twitter/the-algorithm/issues
Twitter's Recommendation Algorithm
51–60 of 1001 posts
Re: Twitter's Recommendation Algorithm
#52From https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92... ( "author_is_elon", candidate => candidate .getOrElse(AuthorIdFeature, None).contains(candidate.getOrElse(DDGStatsElonFeature, 0L))), ( "author_is_power_user", candidate => candidate .getOrElse(AuthorIdFeature, None) .exists(candidate.getOrElse(DDGStatsVitsFeature, Set.empty[Long]).contains)), ( "author_is_democrat", candidate => candidate .getOrEls…
The author_is_elon flag doesn't surprise me, but the two political designators are somewhat shocking. I'd sure like to know what changes based on what Twitter knows about your political affiliation.
I suspect that these are used for metrics tracking rather than being fed back into the recommendation engine. But there's no real way to know for sure given the limited release. These predicates aren't actually used anywhere in the code that's been made available.
Re: Twitter's Recommendation Algorithm
#53[flagged]
Re: Twitter's Recommendation Algorithm
#54From https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92... ( "author_is_elon", candidate => candidate .getOrElse(AuthorIdFeature, None).contains(candidate.getOrElse(DDGStatsElonFeature, 0L))), ( "author_is_power_user", candidate => candidate .getOrElse(AuthorIdFeature, None) .exists(candidate.getOrElse(DDGStatsVitsFeature, Set.empty[Long]).contains)), ( "author_is_democrat", candidate => candidate .getOrEls…
The author_is_elon flag doesn't surprise me, but the two political designators are somewhat shocking. I'd sure like to know what changes based on what Twitter knows about your political affiliation.
So many questions. How are users tagged D or R? Is that a manual process or automated somehow? What is the effect of these tags? Can I find out if my Twitter account is in one of those buckets?
Re: Twitter's Recommendation Algorithm
#55From https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92... ( "author_is_elon", candidate => candidate .getOrElse(AuthorIdFeature, None).contains(candidate.getOrElse(DDGStatsElonFeature, 0L))), ( "author_is_power_user", candidate => candidate .getOrElse(AuthorIdFeature, None) .exists(candidate.getOrElse(DDGStatsVitsFeature, Set.empty[Long]).contains)), ( "author_is_democrat", candidate => candidate .getOrEls…
Only used for metrics, apparently. [0]
/**
* These author ID lists are used purely for metrics collection. We track how often we are
* serving Tweets from these authors and how often their tweets are being impressed by users.
* This helps us validate in our A/B experimentation platform that we do not ship changes
* that negatively impacts one group over others.
*/
Re: Twitter's Recommendation Algorithm
#56Re: Twitter's Recommendation Algorithm
#57Does it show the part where is recommends Elon more than anyone else?
I couldn't find anything specific to that, but I did find thus blurb where they seem to explicitly track how often they're serving Elon's tweets for A/B testing experiments: https://github.com/twitter/the-algorithm/blob/7f90d0ca342b92...
Re: Twitter's Recommendation Algorithm
#58I'm not opposed to social media feeds having complex recommendation algorithms. I just wish they allowed you to opt in to a reverse chronological feed of only people you follow, like RSS.
Twitter has this now. The home page is split into two tabs: "For you", the algorithmic feed, and "Following", the reverse chronological feed of just who you follow.
On my “following” tab (on the phone app) , I’m still getting recommendations for bomb throwers I don’t follow. Am I weird? It’s like an unhinged relative. Not pleasant.
Edit: I reversed “for you” and “following” in my original reply.
Re: Twitter's Recommendation Algorithm
#59the-algorithm is such a pretentious name for a repo
It's a colloquial term for recommendation engines, how often do you hear people say "the algorithm" (vs. "the recommendation engine") on YouTube?
Re: Twitter's Recommendation Algorithm
#60Context: I teach at Princeton and study social media and recommendation systems. From a very quick skim of the repositories, this appears to be quite limited transparency. The documentation gives a decent high-level overview of how Tweet recommendation works—no surprises—and the code tracks that roadmap. Those are meaningful positive steps. But the underlying policies and models are almost entirely missing (there are…
Did you also skim the accompanying (or rather, main) repo, https://github.com/twitter/the-algorithm ?
From a quick clone and line-count, it has:
235 kLOC .scala
136 kLOC .java
22 kLOC .py
7 kLOC .rs
I also haven't skimmed this code except very superficially, but perhaps you should since you're out there making statements with your Princeton credentials.
(I posted this comment with the heads-up a few minutes after your comment above and then expanded it as you didn't respond.)