[Moderators, I apologize if this thread qualifies as spam, and understand if you delete it.]
I could really use your help! If you have a few minutes to spare, and/or you happen to like jokes, read on
In the next *week or two* (and the sooner the better), I need to collect *as much data as possible* from Jester, a joke recommender system that I've co-developed at Berkeley. You rate a few jokes, and then it recommends jokes to you that you'll hopefully enjoy.
- If you could please use Jester (*no registration required*), I'd really, really appreciate that! It's fun: you just read and rate jokes, and you keep going for as long as you want (but the longer you go, the more it'll help)
- If you could tell your friends about Jester, that would be even more awesome.
- You could digg my digg.com post: Digg - Jester: The Online Joke Recommender Just Got Smarter
- You could rate my reddit.com post: Jester: The Online Joke Recommender Now Uses a New Algorithm, Eigentaste 5.0 (reddit.com)
Jester is available at http://eigentaste.berkeley.edu
*Important Note*: Please be honest with your ratings! I need *accurate* ratings data--I don't have any personal connection to the jokes themselves.
If you'd like to know more about my research (and why I need to collect data), read on:
Since last summer ('06), I have been doing research on recommender systems (Recommender system - Wikipedia, the free encyclopedia
) at Berkeley's Automation Sciences Lab under Professor Ken Goldberg, specifically on Jester (and the algorithm behind it, Eigentaste). I wrote Jester 4.0, the latest version of Jester, last summer.
I just recently changed Jester's recommendation algorithm from Eigentaste to Eigentaste 5.0, an algorithm that I developed this summer, together with a graduate student and Professor Goldberg. The graduate student and I are presenting our paper that outlines this new algorithm at a conference in mid-October, and we need statistically significant numbers to discuss at the conference. That's where you come in.