[Wikipedia-l] Re: trust metrics

Mon Feb 16 04:46:35 UTC 2004

Elisabeth Bauer wrote:

> Another idea: a system where you can rate users, but the results are 
> not shown directly. Instead, a score for the articles a user worked on 
> is calculated. the results are printed on the article page, such that 
> you can see: on this article worked people with a very high (or low) 
> rating.

Why not just rate the articles directly, instead of the users? I would 
be much more comfortable with this kind of system. It's less vulnerable 
to the kind of personal animosity that promotes negative feedback wars. 
And suppose you want to consider the relative value of individual 
contributors to an article, as Delirium suggested. If the ratings are 
preserved and tracked on the page history, you could see whether a 
user's contributions generally improved the ratings or not. That's 
actually better information than you could get from giving ratings to 
users. A user could have a great rating, but not necessarily be the best 
informed on that particular subject.

It could be like some companies set up help pages, something along the 
lines of, "Please rate this article: How useful was the information to 
you?" Has the idea of rating articles been discussed before? It could 
also provide additional information for selecting featured articles.

--Michael Snow