Tue 8 Apr 2008
Riffing off some of my earlier posts (namely, Getting SemWebTech to Take, Opportunity Knocks, Community Semantics), I’ve been thinking a lot about the dynamics of meta data generation. I keep mentioning the value of latching onto activities in which people are already engaged, in order to devise processes that “catch on” or can be self-fulfilling. An aspect I haven’t talked so much about is - where in the processes the meta data gets created and captured.
In one context, the “service”, a proprietary hosted platform enables interconnection and linkages, profile-generation through the making of those associations seems the standard approach - whether the profile lives with the host (generally the case), or with the user (rarer).
There’s another context that is particularly interesting, and that is the “tool” approach, which I like to think of as an out-of-body (meta-body?) experience. Here, the mechanism is external to the platform or primary service being utilized, but is a mashable resource that can be tapped for identification and enablement of association with relevant content - and which can generate, on-the-fly, meta data which can then be cross-utilized, and reliably integrated into (or or leveraged for) further activities.
Zemanta and BlueOrganizer strike me as philosophically aligned with the latter, and Alt recently posted some ideas along these lines as well.
April 8th, 2008 at 11:25 am
Hi from Zemanta!
I’d be wondering what kind of metadata would you find useful to be suggested by Zemanta service?
Something that would bring value to writer and/or reader of the content?
bye
andraž
April 8th, 2008 at 12:25 pm
My understanding is that Zemanta iteratively suggests tags and references that a blogger might consider to be relevant/useful.
April 8th, 2008 at 1:08 pm
you are right.
tags, references, images and links
i’d be interested what else writer would like to include in his own articles - because of him or because of his readers?
April 8th, 2008 at 7:19 pm
That’s the million dollar (er, 182 million Tolar) question. My first response would be to enable leveraging of the contexts that authors and readers latch onto (i.e. those they validate through selection), and not just when they are in edit mode. The other would be to go outside of blogging activities.
April 9th, 2008 at 8:21 am
Ah, improvement through feedback - be it real-time or learning on gathered dataset is a must, I agree.
The question is more can some other automatism be offered that we overlooked?
bye
andraz
April 29th, 2008 at 3:04 pm
BlueOrganizer strikes me as a limited tool pegged to specific commerce venues. Does it do much more than the original author could by just hyperlinking to Amazon? In short, is it just a glamorized Google search? A more up-front kind of spyware?