Comments on: The Napoleon Dynamite Problem

By: ★ Technology News | Tech Crown » They Did It! One Team Reports Success in the $1m Netflix Prize

Mon, 29 Jun 2009 11:49:29 +0000

[...] Dynamite, which some people loved and other people hated, get thrown into the mix. It’s nearly impossible to predict whether a person will like films like [...]

By: They Did It! One Team Reports Success in the $1m Netflix Prize | Techdare

They Did It! One Team Reports Success in the $1m Netflix Prize | Techdare — Mon, 29 Jun 2009 05:11:14 +0000

[...] Dynamite, which some people loved and other people hated, get thrown into the mix. It’s nearly impossible to predict whether a person will like films like [...]

By: 大功告成！Netflix大奖得主浮出水面 « 每日IT新闻，最新IT资讯，聚合多站点消息，保证你与世界同步

大功告成！Netflix大奖得主浮出水面 « 每日IT新闻，最新IT资讯，聚合多站点消息，保证你与世界同步 — Mon, 29 Jun 2009 00:00:34 +0000

[...] 像《大人物拿破仑》这样人们爱恨参半的电影，情况是很难判断的。机器几乎无从判断一个人会不会喜爱这部电影（link）。 [...]

By: 大功告成！Netflix大奖得主浮出水面 - 读写网唯一官方中文站 - 搜狐IT独立群体博客

大功告成！Netflix大奖得主浮出水面 - 读写网唯一官方中文站 - 搜狐IT独立群体博客 — Sun, 28 Jun 2009 23:43:13 +0000

[...] 像《大人物拿破仑》这样人们爱恨参半的电影，情况是很难判断的。机器几乎无从判断一个人会不会喜爱这部电影（link）。 [...]

By: BiTTechnology » They Did It! One Team Reports Success in the $1m Netflix Prize

Sat, 27 Jun 2009 18:24:16 +0000

[...] Dynamite, which some people loved and other people hated, get thrown into the mix. It’s nearly impossible to predict whether a person will like films like [...]

By: They Did It! One Team Reports Success in the $1m Netflix Prize | eMediaOne

They Did It! One Team Reports Success in the $1m Netflix Prize | eMediaOne — Sat, 27 Jun 2009 13:32:34 +0000

[...] Dynamite, which some people loved and other people hated, get thrown into the mix. It’s nearly impossible to predict whether a person will like films like [...]

By: They Did It! One Team Reports Success in the $1m Netflix Prize | google android os blog

Sat, 27 Jun 2009 12:15:00 +0000

[...] Dynamite, which some people loved and other people hated, get thrown into the mix. It’s nearly impossible to predict whether a person will like films like [...]

By: Transparency or FAIL | The Noisy Channel

Transparency or FAIL | The Noisy Channel — Wed, 15 Apr 2009 19:33:22 +0000

[...] long been proponent of transparency in search engines and recommendation systems, on the grounds that transparency cultivates trust even in the face of the inevitable fallibility [...]

By: How Recommendation Engines Quash Diversity | The Noisy Channel

How Recommendation Engines Quash Diversity | The Noisy Channel — Wed, 25 Feb 2009 02:16:03 +0000

[...] regular readers here know, I have strong opinions about how recommendation engines should work. So does Daniel Lemire, a regular reader who specifically argues in favor of diversity in [...]

By: Daniel Lemire on Diversity in Recommender Systems | The Noisy Channel

Daniel Lemire on Diversity in Recommender Systems | The Noisy Channel — Mon, 24 Nov 2008 21:42:01 +0000

[...] As regular readers know, I’m also in favor of diversity in recommender systems, but I’m more concerned with their transparency. [...]

By: Daniel Tunkelang

Daniel Tunkelang — Fri, 21 Nov 2008 23:33:31 +0000

As I replied on Twitter, I think your disagreement is that I don’t take a strong enough position against the value of single-metric recommendation systems. Which I interpret as your agreeing with me, only more so.

By: Recommender systems: where are we headed?

Recommender systems: where are we headed? — Fri, 21 Nov 2008 22:57:49 +0000

[...] Tunkelang comments on the recent progress in collaborative filtering: (…) the machine learning community, much [...]

By: Daniel Lemire

Daniel Lemire — Fri, 21 Nov 2008 22:37:53 +0000

First of all, I disagree that even if this is one-shot recommendation, they are on the right track, because I disagree that accuracy is all that matters, see:

http://www.daniel-lemire.com/blog/archives/2008/11/14/measuring-the-diversity-of-recommended-lists-at-last/

Hint: in IR, nobody would focus entirely on precision at the expense of recall. We know that a balance is needed. Yet, in collaborative filtering, people use a single metric, without any balance.

But even so, is their accuracy likely to pan out in the real world? Take into account that they work with static data set… ignoring the feedback effect:

http://www.daniel-lemire.com/blog/archives/2007/12/22/collaborative-filtering-why-working-on-static-data-sets-is-not-enough/

Hint: in practise, users will react to your recommender system and not rate the same items. This may play in your favour or against you.

I need to write a position paper of some kind.