As I’ve mentioned earlier, I read a lot of feeds. By organizing them in categories, I can easily get an idea of what field is hot. For example, yesterday, upon looking at Bloglines, my RSS aggregator, I knew that something important was happening (or not, considering that it was April 1st) in the search space as the feeds I had classified as search-related seem to have more than the usual number of entry. I call this concept blog chatter.
Over the past few weeks, with the 9/11 commission running its public interviews, the word chatter has come back to the forefront. Basically, it means that the noise is increasing. Thanks to the immediate or almost immediate updates given by RSS feeds, it is now possible to track what the blogosphere is talking about. Of course, this is hardly news to users of Technorati, Blogdex, Popdex, Daypop or similar services. However, one thing that is missing is a graphical representation if this chatter, similar to what Newsmap used to do.
Newsfilter does an interesting job of creating instant layout based on news chatter in the Google News service but I’ve yet to see anything similar for the blog world. This is where I go back to the subject of categorization in RSS feeds. If we could define some agreed upon level of categorization (or the skeleton of one) and/or dictionary of synonyms in the category space, we could start creating broad baskets of feeds. Feedster is trying to do some of that grouping with their concept of Feedpapers but I would like to question whether this task has to be a manual one.
If there were an agreement between blog authors as to some standard conventions in terms of defining a basic categorization scheme (or using an existing one: after all, there are a number of categorizations schemes available, some privately held (like the Dewey Decimal System), and others in the public domain (like the Library of Congress MARC format). I am not advocating one over another but I do thing that this type of organization could be useful.
Once feeds are more organized, one could generate applications that would pump out not just list of links that were popular but also provide visual mappings of what’s important in the blog world. I know it sounds far-fetched but similar attempts at information organization were made in the late 1990s. Considering Moore’s law, we should now have the processing power for those more processor-intensive applications to run. The idea here is to be able to get a set of visual cues relating to what subjects are hot or not. Looking into a 3D space, one could map number of individual mentions on one axis, number of sites mention on another, and increase or decrease on a third one. Color could also be used to indicate the speed of increase or decrease in the mentions. All this would allow to get a quick glance at what “chatter” is occurring. Clicking on a resource would then expose richer data. It seems all the pieces are there to build such an application. My knowledge of 3D interface design is limited, however, so I can’t build it myself. Anyone out there cares to do it?