Seb's Open Research
Pointers and thoughts on the evolution of knowledge sharing
and social software, collected by Sébastien Paquet

email me

My keywords
My popular pieces
Stories and articles
2002 weekly archives
2003 weekly archives
2004 weekly archives
Neighborhood tour
Technorati cosmos
Blogstreet profile
Today's referers
Seb's home

My other weblogs:
Many-to-Many: Social Software groupblog
My public mailbox
My 'Quantum Bits' blog
En français SVP!

Montreal, QC
Social Software
Augmented Social Net
Emergent Democracy
New webloggers

Research Blogs
Minciu Sodas
First Monday
AmSci E-Prints
Weblog Kitchen
Knowledge Board
Weblogs at Harvard
EduBlogging Network
Reputations Research
Know-How Wiki

JC Perreault
Smart Mobs
Blue Oxen
Web Tools- Learning

with a weblog

Spike Hall
Chris Dent
John Baez
Bill Tozier
Erik Duval
Clay Shirky
Jill Walker
Jim McGee
David Tosh
danah boyd
Sylvie Noël
John Taylor

Ton Zijlstra
Joseph Hart
Ed Bilodeau
Peter Suber
David Deutsch
David Brake
Steve Cayzer
Lilia Efimova
Mark Hemphill
Alex Halavais
Mike Axelrod
Paul Resnick
Cosma Shalizi
Andrew Odlyzko
Lance Fortnow
Tom Munnecke
Henk Ellermann
Mark Bernstein
Jeremy Hiebert
Jacques Distler
Michael Nielsen
Thomas N. Burg
Hassan Masum
Ian Glendinning
Marc Eisenstadt
George Siemens
Howard Rheingold
Stephen Downes
John Bethencourt
Sebastian Fiedler
Kevin Schofield
José Luis Orihuela
Martin Terre Blanche
Elizabeth Lane Lawley
Paul Cox
Jon Udell
Don Park
*Alf Eaton
Lion Kimbro
Phil Wolff
Jay Cross
Julian Elvé
Matt Webb
Adina Levin
*Marc Canter
Matt Mower
Kevin Kelly
Dina Mehta
Greg Searle
Ross Dawson
Al Delgado
Rajesh Jain
Lee Bryant
Jesse Hirsh
David Sifry
Jeff Bridges
Stowe Boyd
Walter Chaw
Piers Young
Barbara Ray
Dave Pollard
Ian McKellen
Josep Cavallé
Hylton Jolliffe
Lucas Gonze
Jerry Michalski
Chris Corrigan
Boris Anthony
Michael Fagan
Mary Messall
Denham Grey
*Ross Mayfield
*Phillip Pearson
Whiskey River
David Gurteen
Tom Portante
Chris Wenham
Pierre Omidyar
Stuart Henshall
Greg Costikyan
David Gammel
Renee Hopkins

Peter Van Dijk
Peter Lindberg
Michael Balzary
Steven Johnson
Robert Paterson
Eugene Eric Kim
Jason Lefkowitz
*Flemming Funch
Bernie DeKoven
Edward De Bono
Maciej Ceglowski
Charles Cameron
Christopher Allen
*Philippe Beaudoin
Richard MacManus
The Homeless Guy
Ward Cunningham
Hossein Derakhshan
Stewart Butterfield
Stefano Mazzocchi
Evan Henshaw-Plath
Gary Lawrence Murphy
Karl Dubost
*Dolores Tam
Norbert Viau
Patrick Plante
Daniel Lemay
Sylvain Carle
Bertrand Paquet - Hydro-Québec
Michel Dumais
Mario Asselin
Robert Grégoire
Roberto Gauvin
Clément Laberge
Stéphane Allaire
Gilles Beauchamp
Jean-Luc Raymond
without a weblog
Steve Lawrence
Simon B. Shum
Stevan Harnad
Brian Martin
John Suler
Christopher Alexander
Johanne Saint-Charles
Douglas Hofstadter
John Seely Brown
Murray Gell-Mann
Steve Newcomb
Howard Gardner
Anthony Judge
Patrick Lambe
Donald Knuth
Phil Agre
Jim Pitman
Chris Kimble
Peter Russell
Roger Schank
Howard Bloom
John McCarthy
John C. Thomas
Doug Engelbart
Seymour Papert
Hossein Arsham
W. Brian Arthur
N. David Mermin
Tommaso Toffoli
Brian Eno
Will Wright
Jean Leloup
Daniel Boucher
Daniel Bélanger
Laurence J. Peter
Plume Latraverse
George Pólya
Thomas Kuhn
Edsger Dijkstra
Hermann Hesse
Abraham Maslow
Benjamin Franklin
Shiyali Ranganathan
Andrey Kolmogorov
Jiddu Krishnamurti
Georges Brassens
Bertrand Russell
Astor Piazzolla
Kurt Cobain

Google Search
Fagan Finder Blogs

Google Glossary

NEC ResearchIndex e-prints
SEP Bibliography
citebase search

Complexity Digest
Principia Cybernetica

All Consuming
gnod musicmap
Logical Fallacies
W3C Link Checker
Wayback Machine
RemindMe Service

Music streams:
Radio Tango Argentino
Boombastic Radio
Limbik Frequencies
Radio Paradise
shoutcast streams
electronic streams index


Subscribe with Bloglines

Subscribe to "Seb's Open Research" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.



Towards structured blogging

(French language translation: Vers un carnet Web structuré - merci, Dolores!)


Lately I've been thinking about how we could evolve blogging tools to allow people to author more structured (dare I say semantic?) content, so that other people could find their stuff that they find of interest more easily.

Right now what we have, globally speaking, is pretty much a huge pool of blog posts, each implicitly tied to a particular weblog author and with a date slapped on. Now, say I've written a review of the latest Radiohead album into my blog. I'd like others who are interested in Radiohead, or in music reviews in general, and who may not know me, to be able to pick out my review from the common pool in a simple way. Interesting people may come my way because of this.

What we're talking about is getting people to put more metadata on their content. Now allowing it is one thing, and fostering it is another. And I'd say the latter is the bigger challenge.


The importance of feedback

I believe a critical element to get a sustainable system is for people to get reasonably quick feedback in return for the extra effort expended in creating metadata. The Internet Topic Exchange, in which Phillip Pearson implemented the ridiculously easy group-forming design, seems to work because it has a short feedback loop, and might provide a template for where we'd like to go. Here's an illustration of how the Internet Topic Exchange works:

People (on the left) associate select posts with particular topics by specifying a Ping URL; TrackBack carries information about their posts into the Exchange (the fat square), and from there they make it into open topic-specific blogs (on the right). (I don't need to talk about the per-topic Wiki pages here.) Creating a new topic is as easy as inventing a name - anyone is allowed (and encouraged) to do it.

Now, people interested in particular topics will watch them, and this is where the fun comes from. Look at the figure above. Say John (the yellow guy) doesn't know Elaine (the purple girl). But he watches Topic 1 dutifully. Now Elaine happens to come across the channel for Topic 1 and posts to it because she's also interested in Topic 1. Soon John gets her post in his aggregator, checks out her blog, and voilà, he's found a like mind.

In this case the metadata that we have managed to get John, Elaine and others interested in consists of a simple topic identification. This simple scheme has helped such groups as spanish bloggers, emergent democrats, and Austin bloggers coalesce.


Beyond topics

How might we work our way from this model to a different form of metadata, say a music review? Here's a possibility. We retain the basic architecture of the Exchange, but add a new type of blog post called "music review". A drop-down box might enable me to select between "plain-vanilla blog post" and "Music Review". The interface for entering a "Music Review" post might look like this:

(Here I'm assuming there is some sort of standard for music reviews composed of a title, an identification code for the piece of work being reviewed, a rating, and some text. If you know of a similar existing standard please let me know.)

In an ideal world, the "Find" button would pop up an assistant that lets you dig into a music metadatabase (say, MusicBrainz) and quickly home in on an unambiguous ID code for whatever it is you're reviewing (possibly asking you to contribute to the metadatabase if it doesn't know about this piece of work), and puts it in the box.

You fill in the other boxes, click "Post", and your part is pretty much done. Next the system does three things. First, it stores the review in a non-lossy format somewhere on your site. Second, it converts your input into a regular blog post to put in your blog and RSS feed. Third, it notifies one (or more) central indexing service(s), analogous to the Topic Exchange, that knows about music reviews, of the availability of your new review and of its location. This central service also serves a variety of RSS feeds. There could be song-based feeds, album-based feeds, artist-based feeds, genre-based feeds, etc. that you can subscribe to. You could subscribe to the "Radiohead songs and albums" feed and get all reviews of Radiohead songs and albums as soon as they come out. Or maybe you just want to be notified whenever someone reviews a certain song you especially like.

As in the case of the Topic Exchange, these RSS feeds are where the feedback (and the addictive quality) comes from and how new interpersonal links form and people cluster.


Putting it all together

Now to generalize. I talked about music reviews, but the scheme should also work with other kinds of content. Music video reviews, movie reviews (using IMDB?), ad reviews, TV show reviews, game reviews, radio station reviews, weblog reviews, restaurant reviews (perhaps using GPS data), scholarly article reviews... other kinds of content that are not reviews also, such as song lyrics, TV show transcripts, quotes, self-identification data.

Each of these would have (one or more) standard format(s). The basic idea is to let the blogging system support "alternate post type" plug-ins. Anyone should be able to develop such a plug-in (and a corresponding indexing service) for a new kind of post that they want.

If something like this were to become successful it would compete with a host of commercially led user-contributed databases such as Amazon's review database. One advantage would be to put control more firmly in the hands of contributors.

Will people care enough about having their writings under their own control and collected in one place to move away from such databases? I think they might. I know I do.

Other people are thinking along similar lines: Danny Ayers, Alf EatonMarc Canter, Sam Ruby, Stefan Smalla, Ben Trott, Karl Dubost, to name a few. It would be nice to make something like this happen in 2003.

What do you think? [7] 

Backlinks via Technorati


Click here to visit the Radio UserLand website. Copyleft 2006 Sebastien Paquet.
Last update: 4/22/2006; 12:27:06 PM.
This theme is based on the SoundWaves (blue) Manila theme.