I think this puppy warrants wider dissemination. Working with Sam Hunting
on a Weblog federation scheme, I have made just a bit of progress in the
direction of thinking through what it takes to *really* federate, not just
aggregate, RSS feeds from blogs and other sources. I have some ideas and a
small demonstration. Slides can be viewed at http://www.nexist.org/rss/slides/ (01)
That work is being done on behalf of Tom Munnecke's Uplift Academy project.
Federation is part of a mechanism being invented by which uplift patterns
can be mined or otherwise discovered, e.g., by collaborative filtering,
from documents of all sorts found on the Web. It is thought that simple
aggregation of blog content does not necessarily contribute to the
reduction of information overload, though aggregation does, indeed, serve
the important purpose of bringing together all the content obtainable by
subscribing to RSS feeds. The hypothesis behind my project holds that by
aggregating RSS feeds into one well-formed topic map, all of the many
benefits of topic maps, including their indexical and associative powers,
can contribute, at once, to a significant reduction of info-glut or
whatever you want to call it, and to a more powerful, emergent, sense of
community and affinity among those who participate. (02)
Meanwhile, in all of my vast and expansive roaming the landscape covered by
google, what follows is information I somehow missed. (03)
Requisite quote:
"How many bloggers does it take to screw in a lightbulb?" (04)
Enjoy.
Jack (05)
>The folks at Blog Talk have really put together a good show (more at
>http://blogtalk.net/panelists.html ). I was impressed the number of
>women participants in this panel, as well as diversity of people --
>new faces! Well, compared to other conferences. Our old friend Phil
>Wolff is speaking.
>
>And then there's HP's Paul Shabajee and Steve Cayzer's semantic blog
>effort, which is going to be interesting to watch, (found at
>http://jena.hpl.hp.com:3030/blojsom-hp/blog/ and position paper at
>http://www-uk.hpl.hp.com/people/steve_cayzer/papers/scps-submitted-
>030228.htm).
>
>What I particularly like about Paul and Steve's effort is that they
>talk about semantic blogging, the semantic web really, from two
>different angles. There is the smart metadata that's made accessible
>thought RDF, such as the efforts with RSS, true. But they also talk
>about the ontologies, though the paper focuses more on classification
>than concepts, which is more my thing.
>
>The first angle of what they cover has been the complete focus so far
>on this discussion group -- ThreadsML, RSS, Trackback and so on. If I
>push back at RSS, it's because for the last month I've seen RSS and
>semantic combined, but with no explanation as to why there's a
>semantic element to RSS -- being recorded in RDF does NOT make a
>specification 'semantic'. Some of the semantic qualities of RSS (and
>FOAF and other non-ontological specifications) started to come out
>here, but then the conversation tends to drill down to a pure tech
>level quickly, rather than stay at a level that will invite
>discussion with less technical folks in the audience.
>
>The second angle of the HP talk is the one that interests me the
>most -- the ontologies. Unfortunately, in the talk, this seems to be
>limited primarily to classification and smart linking as well as
>discovery across weblogs. However, to be expected with a 'semantic
>blog' talk.
>
>But ontologies can be richer than this, doing this such as allowing
>us to search on 'concepts' as well as keywords
>
>Regardless, a couple of things from the talk: I think this talk could
>point out something fairly important -- consistency of metadata. I
>know from a threading perspective, there's trackback, ThreadsML, and
>David's work at Blogmatrix (other efforts, too?) However, as someone
>just pointed out in a trackback at my weblog, trackback's 'too hard'.
>If TB's too hard, what about the more complex specs? And how are we
>going to have rich connectivity if everyone does their own thing?
>
>Trackback has wide spread implementation -- it's too bad that the
>ThreadsML folks et al aren't building on this. It's RDF -- can always
>add namespace and extend.
>
>Second is the necessity of ontologies if we're ever going to break
>out of the keyword jail. But that's for another day.
>
>
>
>
>
>------------------------ Yahoo! Groups Sponsor ---------------------~-->
>Rent DVDs from home.
>Over 14,500 titles. Free Shipping
>& No Late Fees. Try Netflix for FREE!
>http://us.click.yahoo.com/BVVfoB/hP.FAA/uetFAA/qkHolB/TM
>---------------------------------------------------------------------~->
>
>To unsubscribe from this group, send an email to:
>blogunlimited-unsubscribe@yahoogroups.com
>
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ (06)
---------------------------------------------------------------------------
XML Topic Maps: Creating and Using Topic Maps for the Web.
Addison-Wesley. Jack Park, Editor. Sam Hunting, Technical Editor (07)
Build smarter kids globally to reduce the need for smarter bombs. (08)
--
This message is archived at: (09)
http://collab.blueoxen.net/forums/cgi-bin/mesg.cgi?a=tools-yak&i=5.1.1.6.0.20030509071404.02b4b418@thinkalong.com (010)
|