TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Recommending items to more than a billion people

116 pointsby antoinecalmost 10 years ago

6 comments

sjtrnyalmost 10 years ago
The technical name for this is &quot;collaborative filtering&quot;. I think they are basing their work on this paper<p>- <a href="http:&#x2F;&#x2F;www.jmlr.org&#x2F;papers&#x2F;volume10&#x2F;takacs09a&#x2F;takacs09a.pdf" rel="nofollow">http:&#x2F;&#x2F;www.jmlr.org&#x2F;papers&#x2F;volume10&#x2F;takacs09a&#x2F;takacs09a.pdf</a><p>EDIT: Actually looks like Eq (15) from<p>- <a href="http:&#x2F;&#x2F;public.research.att.com&#x2F;~volinsky&#x2F;netflix&#x2F;BellKorICDM07.pdf" rel="nofollow">http:&#x2F;&#x2F;public.research.att.com&#x2F;~volinsky&#x2F;netflix&#x2F;BellKorICDM...</a><p>Anyway there are lots of papers around on the topic.
评论 #9651412 未加载
评论 #9652763 未加载
istvan__almost 10 years ago
This is pretty cool, the scale is one reason almost any time Facebook publishes something in &quot;big data&quot; subject it is worth to read.
评论 #9652642 未加载
a1k0nalmost 10 years ago
FWIW, I gave a talk about the Alternating Least Squares algorithm mentioned here (and linked in several comments) and how we implemented it at Spotify:<p>Slides: <a href="http:&#x2F;&#x2F;www.a1k0n.net&#x2F;spotify&#x2F;ml-madison&#x2F;" rel="nofollow">http:&#x2F;&#x2F;www.a1k0n.net&#x2F;spotify&#x2F;ml-madison&#x2F;</a> Video (for the extremely patient): <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=MX_ARH-KoDg" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=MX_ARH-KoDg</a>
acconstaalmost 10 years ago
<i>To solve the matrix equation A × X = B we need to find the inverse A^-1</i><p>Huh? Isn&#x27;t Gaussian elimination more straightforward?
评论 #9650973 未加载
评论 #9651018 未加载
评论 #9650782 未加载
评论 #9651037 未加载
FiReaNG3Lalmost 10 years ago
I hoped for a minute that they shared their complete implementation; anyone aware of a recommendation system that can scale to millions of items, be updated as soon as new items come in (no full graph recalculation) and take multiple inputs (ratings, saved in library, etc)?
评论 #9651586 未加载
skbohra123almost 10 years ago
Please don&#x27;t do it, however great technical feat it is, the truth is, it sucks. I hate those the most in facebook.
评论 #9651043 未加载