Have been exchanging ideas on blog categorisation with a number of people, and got me hunting again for what other folk are up to out there on the internet to combat the noise-to-signal ratio. Of course my interests lie in adding value to the Goog content I aggregate.
This is less about dehumanising aggregated blog posts into a purely technical stream of content and more about adding value to the content itself. It's great to give voice to all the MX blogs in one place -- especially for those infrequent bloggers who might otherwise go overlooked -- but as the volume of posts go up the capacity for visitors to effectively digest what is going on goes down.
Personally, I look to the Goog for more than just the latest MX techno tip -- it's like taking the pulse on the Macromedia community -- seeing what folks are interested in, what's upsetting them, what the future holds. But still -- there is tremendous value in a resource that provides a conduit through to real technical experience.
So what we need is something that allows us to focus in on specific topics when the need arises -- to go beyond a rolling newsticker a repository that has archival value. This means better categorisation at least. But there are problems...
Auto trackback and categorising blogs
"There are two problems to overcome:
* People are notioriously bad at categorising things accurately or consistently
* Predefined controlled vocabularies ensure consistency but laziness or uncertainty by authors means these CVs are rarely used well, unless financial incentives are attached."
Azeem is spot on here. And I've remarked on simlar issues Fullasagoog has encountered.
Azeem goes onto suggest the use of public taxonomies like DMOZ -- I was hoping the Mothership might have something similar. I tracked back through Azeem's post to more discussion on taxonomy and the possiblity of an automated service to provide dynamic categorisation.
Categorisation Server
"That leads to some serious thoughts on a taxonomy for the Semantic Web. A service which provided automatic categorisation based on a number of existing taxonomies, say DMOZ as Azeem suggests plus ISO, W3C, etc. etc., would be very useful. I think a Web Service application that fulfilled such an application would be ideal and would mean integration with existing tools would be possible easily."
Perhaps Google's acquistion of Blogger might lead to such a thing?? But I think I'm going to have to come up with something before then :)
More on categorising blogs
"These would have a number of requirements:
* creation of many, multi-faceted taxonomies--to ensure they are decentralised, vendor-independent. These taxonomies would need to be dynamic, in the sense that reconfiguration should be a property of the system which creates them
* a way of looking at a blog posting (with its cues and clues) and spitting out a set of suggested categories, using some neat technology."
"how we can create a market for taxonomy definition allowing multiple taxonomies to emerge, what cues you could use to generate category (and "others who might like to see this" information)"
Again most of the issues of categorisation stem from the need to come up with a half decent -- yet ever evolving -- hierarchy of categories. Then of course you've got to manage the thing and make it dead simple for people to use. Bah.. I might as well be teaching my Grandmother Java -- it would be easier :)
Scrambled Googs"
"I don't know anything about Full As A Goog's underlying architecture, but I've got some suggestions for Geoff..."
Roger comes up with some good ideas -- of course there are already many people attempting to demystify categorisation, I just need to look harder.
I've been poking at XFML for distributing a simple blog roll of the Goog's Blogs On Tap The XFML Core is an open XML format for publishing and sharing hierarchical faceted metadata and indexing efforts.
Roger suggests the ENT extension to RSS. The ENT specification describes a simple format for extending the RSS 2.0 syndication format to include topical information. I like it enough to want to have a play -- read it works for my simple brain -- so I'll come back with more when I'm done.
Posted by modius at 03:55 PM | Permalink
Trackback: http://blog.daemon.com.au/cgi-bin/dmblog/mt-tb.cgi/108
I think a big issue that involves information design and useability is that these charts are often dual purpose.
Posted by: bali on December 14, 2003 05:56 AM