Mapping the blogosphere

[Note : this is the sequel to the post published this Monday(2007/04/23) ]

I mentioned in Monday's post that blogosphere has became the biggest social network ever seen. Photo and story courtesy of Discover Magazine available in Discover's web site. Some of blog posts point to the article. If you're new to this topic, the Blogosphere Mapping Gallery may be of interest. Eventually, the map was built to visualize the interactions and relations between assorted of webs and blogs. It's a typical application of visualization.

I couldn't manage to find the description of clustering algorithm used by the creator of the map.I'll keep on following this topic in the upcoming weeks.

The creator of the map,Mathew Hurst, the co-creator of Blogpulse, said the most interesting post to date is "DailyKos Featured in Discover Magazine". Studying the differences in the structures of the blogosphere has suggested that bloggers in the political arena have denser linking behaviours. Connectivity is fundamental to politics. Consequently, it is not surprising to see a natural interest in understanding the nature of discourse.

(這幾天都在忙這些有的沒的,該看的書也沒有看,罪惡感越來越重,不過憑良心講,Visualization 是真的很有趣)

Comments

  1. I think it's a force-based algorithm working over the surface of a sphere. He's got an interesting way of doing clustering also: http://datamining.typepad.com/data_mining/2006/09/analytical_grap.html

    ReplyDelete
  2. Thanks for your information. I'd downloaded the paper mentioned in the post. It'd been added to my reading list. :)

    ReplyDelete

Post a Comment

Popular posts from this blog

如何評估推薦系統(一)

備忘:在 Mac OS 環境安裝 Python 開發工具

[詩戀] 鄭愁予/定