I group social software into two broad categories--active, and latent. Active social software actually supports social interaction, eg.g Meetup, Hydra, Uncle Roy, IAwiki. Latent social software is software that derives value from social systems, and makes the results available to third parties--blogdex, Technorati, historyflow. The Waypath Buzz Maker is latent social software that indexes weblogs and graphs up to 5 word frequencies against each other. Here, for example, is the graph for the words Friendster, Fakester, and Tribe.net: Unfortunately, as you can see, the data is nearly worthless. It suffers from several classic mistakes -- the database was too small in the beginning to be worth using, collapsed again in August, and is being added to cumulatively, so all searches also collapse at the end. In addition, the word count is not indexed to overall posts, so even if occurrences of a word are falling in relative terms, it can still be presented as rising if its absolute numbers are going up. The thing that interests me most about Buzz Maker is not the data from it -- the site's creators may or may not ever fix the methodological problems -- but rather the fact that there are a number of such attempts out there to visualize data from social systems (as with Ben Discoe's Friendster graphs which I posted about earlier.) Like recent work on social networking protocols and services, collaborative groupware, and meme and word-burst tracking, the engineering required for this type of latent social software has become simple enough that a small number of people can accomplish a lot in a short time. Even if the Buzz Machine never gets good, its a harbinger of a lot more social visualization tools showing up in the next couple of years.