Lots of interesting posts on folksonomies in the many2many group blog.

Clay Shirky offers an interesting metaphor for how new ideas and technologies, folksonomies in this case, evolve and are adapted.

To put this metaphorically, we are not driving a car, with gas, brakes, reverse and a lot of choice as to route. We are steering a kayak, pushed rapidly and monotonically down a route determined by the environment. We have a (very small) degree of control over our course in this particular stretch of river, and that control does not extend to being able to reverse, stop, or even significantly alter the direction we’re moving in.

Liz Lawley comments (“it’s the social network, stupid!”) on the need for personalized ranking.

One of the things that I’ve tried to emphasize every time I’ve talked to people involved with search engines is the growing uselessness of ranking algorithms that take the search and linking habits of the whole world into account. I don’t want to know what the average eight-year-old calls an image. I want to know what my friends and colleagues call an image. Or a link. Or a photo.

Flickr and del.icio.us work so well for me not because they aggregate the world’s tags, but because they allow me to aggregate my social network’s tags, links, and photos. I don’t want to see everybody’s links on productivity, but I do want to see Merlin Mann’s. I don’t want to see everybody’s links on blogging, but I do want to see danah’s. I don’t want to see “research” resources from a molecular biologist, but I do want to see them from a sociologist studying online social networks.

How does each of us personalize the ranking algorithms used by information retrieval systems? We can tell Flickr who’s in the group of people whose opinions we value. But do we have to do the same for del.icio.us and technorati and the 87 other sites we visit? An obvious idea is to integrate a trust based approach with a system to aggregate and integrate RDF information on our social network (FOAF) and the objects being searched over. One problem is that the straightforward way to define a ranking algorithm is non-incremental and expensive. Even incremental approximations will be expensive for large collections of things to be ranked. Google can afford to do it for the average web user, but not for each of us. Personalized and topic-based ranking offers many challenges (see An Analytical Comparison of Approaches to Personalizing PageRank for some discussion).

RDF + trust might form the foundation for a good motor for our kayak. We’ll have to see if it’s too big.

Related posts: • BusinessWeek ranks 50 most innovative companies;  • Spings Evolved: Comment Spam goes to Ping Servers;  

Posted on Sunday, January 23rd, 2005 at 11:14 am in categories [Semantic Web, Social]. Comments feed. Leave a response, or trackback from your own site. edit