« Announcing The Blog Cooperative |
Main
|
Reputation for better search relevancy »
FOAF as part of a reputation system
I learned about
FOAF some time ago. An XML-based personal profile that you host on your own site, it sounded kinda neat, but there didn't seem to be any cool applications for it (none that I found, at least). Until now. Enter
David Sifry and
Technorati. Technorati has implemented a
Profile feature that utilizes FOAF. This profile information is intergrated into
Cosmos listings, displaying the bloggers photo next to the person's blog, with a link to the Technorati profile. Cool. So this morning I set up my FOAF, using the
FOAF-A-Matic, and my Technorati profile can be found
here. The profiles don't display much information yet, but I have a feeling more blog and blogger information will be added in the future.
This got me thinking about a potential role of FOAF in online reputation systems. I had thought about this before, but I dismissed it because self-hosting your own reputation statistics would not result in enough trust (because it could be easily altered). But the Technorati implementation takes basic profile information from the FOAF file and combines it with linking information from many other sources (blogs). The result is a trusted authority (Technorati) combining profile (FOAF) and reputation (links) information from a wide array of sources to form a picture of person's reputation. All of the data used to create the reputation measurement is stored in a distributed manner, on many different web servers, controlled by many different people. I think this reputation system model has potential.
By the way, I learned about this new Technorati feature via
Marc Canter. For more FOAF information, read Marc's recent posts on the subject:
Here we go!,
Easy FOAF Notes,
Friends of FOAF, and
Getting FOAFy wit it.
Posted by Mark at July 22, 2003 2:44 PM
| TrackBack
Posted by:
Mark Thristan at July 24, 2003 7:08 AM
Very interesting thoughts - I've been thinking vaguely about how search engines could take advantage of networks of trust (like those set up with FOAF) to deliver the missing aspect of reputation as they deliver their search results. A reputation measure would add a human dimension to the value system created in search engines like Google through PageRank etc. I haven't really thought it through to the nth detail, but it strikes me that if you could rate pages that you trust akin to the way we add favourites in our browsers, and then lock into a network of recommendations based on the people we "trust" when searching, there might be that missing level of "you should read this article because your peer group recommends it as well as it being relevant to the search terms that you entered". Sorry if this is a bit garbled, I wrote a bit more about it on my blog last night.
Posted by:
Mark Carey at July 24, 2003 6:47 PM
Mark, that is very interesting. Search results based (in part) on the reputation of those you trust....I think that certainly has a lot of potential. I have decided to create new thread (blog entry) on the topic: Reputation for better search relevancy
Search Engine Optimization, Blog Design, Movable Type Customization, and more.
Very interesting thoughts - I've been thinking vaguely about how search engines could take advantage of networks of trust (like those set up with FOAF) to deliver the missing aspect of reputation as they deliver their search results. A reputation measure would add a human dimension to the value system created in search engines like Google through PageRank etc. I haven't really thought it through to the nth detail, but it strikes me that if you could rate pages that you trust akin to the way we add favourites in our browsers, and then lock into a network of recommendations based on the people we "trust" when searching, there might be that missing level of "you should read this article because your peer group recommends it as well as it being relevant to the search terms that you entered". Sorry if this is a bit garbled, I wrote a bit more about it on my blog last night.