Businessweek Archives

Tagging: What if a machine gave prompts?


? Sun: Brighter prospects for new hires than Google? |

Main

| BW Story on Podcasting Inspired by Podcast Chats ?

April 02, 2007

Tagging: What if a machine gave prompts?

Stephen Baker

I was sorry to miss the Weblogs and Social Media Conference in Boulder last week. But if you click that site, there are links to all the technical papers presented. Awarded best in show was a paper by researchers at Northwestern on an automatic system to help people pick the best tags for whatever they're posting online. It surveys the existing tags, calculates which are the most effective, and assigns a score to evaluate each tag we choose.

Here's are a few thumbnails from the session by Matthew Hurst. One especially provocative point: He cites Danah Boyd, who says the real danger facing young people online is less from sexual predators and other media fodder than from exploitative marketing machinery. I'm looking forward to the video of her talk, which focused on MySpace, teenage culture, and the transitions between the physical and virtual worlds. Matthew says the video will be online shortly.

Longer summary from Valery.net, who groups all of these technologies and initiatives with the ever controversial tag: Web 3.0.

08:22 AM

research, tagging

Stephen,

My favorite paper at ICWSM was from a group of Brazilian academics whose analysis of a mega-dump of ISP indicated that blog popularity and inbound links don't correlate....that there is a whole lot of social navigation going on. The authors said this was an opportunity for search engines to include other dimensions when producing query results. I believe it's an opportunity for the blogosphere to reevaluate its presumptions about influence and popularity. I'm sure you'd agree that inbound links are one indicator of popularity or influence, but certainly are only one proxy and can, in fact, understate popularity or influence.

Here's my brief summary

You missed a good conference.

Max

Posted by: Max Kalehoff at April 2, 2007 08:44 AM

Stephen,

My favorite paper at ICWSM was from a group of Brazilian academics whose analysis of a mega-dump of ISP data indicated that blog popularity and inbound links don't correlate....that there is a whole lot of social navigation going on. The authors said this was an opportunity for search engines to include other dimensions when producing query results. I believe it's an opportunity for the blogosphere to reevaluate its presumptions about influence and popularity. I'm sure you'd agree that inbound links are one indicator of popularity or influence, but certainly are only one proxy and can, in fact, understate popularity or influence.

Here's my brief summary

http://www.attentionmax.com/blog/2007/03/inbound_links_dont_necessarily_correlate_with_blog_popularity.php

You missed a good conference.

Max

Posted by: Max Kalehoff at April 2, 2007 08:44 AM

You mean like this marketing scenario? - http://blog.rebang.com/?p=577

From the comments it's gotten (mostly on other blogs linking to it), a LOT of people are not going to like it, but I've been reading of recent developments that are moving us closer to it.

Posted by: csven at April 2, 2007 09:20 AM


Toyota's Hydrogen Man
LIMITED-TIME OFFER SUBSCRIBE NOW
 
blog comments powered by Disqus