Thanks, Xavi and Leslie.
Now... I'm listening to Sophie Zelmani's music, and trying hard to find a way 
to watch xavi's videos, as youtube seems banned here :-(

As I mentioned in the end of that demo paper, I'm doing some experimental thing 
recently, to answer a futher question related to Aggregator- "Can we get richer 
information from the aggregated things?". Now the answer is clearly yes!

To put it simply, say we've constructed a Knowledge Base with the information 
aggregated, however, 20% of the artists don't have any mood annotations. The 
experiment shows that, with benefits of both semantic-web technologies and this 
framework, it can predict the mood tag of these unknown artists, with 
acurracies of ~70% (note that a random prediction is just of 20% for 5 mood 
classifications). Plus, the whole prediction process is done in a pretty decent 
way.

I think the result is very interesting and promising. With the flexible and 
well-designed framework of CLAM, I'm eager to see what next we may achieve.

Cheers,
Jun

  ----- Original Message ----- 
  From: Leslie Hawthorn 
  To: Xavier 
  Cc: [email protected] ; clam 
  Sent: Friday, April 24, 2009 12:18 PM
  Subject: Re: [CLAM] CLAM presentation at WWW 2009


  Hi Xavier,


  On Thu, Apr 23, 2009 at 6:15 AM, Xavier <[email protected]> wrote:

    Here you can watch the videos of our presentation of the CLAM Aggregator at 
the WWW 2009 conference that is being held in Madrid this week (they are 3 
videos with some discontinuities ):

    http://www.youtube.com/watch?v=_-c762VK83s
    http://www.youtube.com/watch?v=ww42HLa-07s
    http://www.youtube.com/watch?v=lMLHYDbX-0c

    As explained in the talk, this is mostly the result of Jun's work, which 
started with a Google Summer of Code project. Congrats Jun for such a great 
outcome of your work in CLAM!


  Are these videos released under an open content license, e.g. Creative 
Commons? If so, it'd be great to link to them from here:

  
http://www.teachingopensource.org/index.php/Teaching_Materials_Catalogue#Review_in_process

  Jun, congratulations for your great work on CLAM. What are you doing now?

  Best,
  LH 


  -- 
  Leslie Hawthorn
  http://www.hawthornlandings.org
  http://identi.ca/lh
  http://www.twitter.com/lhawthorn



------------------------------------------------------------------------------


  _______________________________________________
  CLAM mailing list
  [email protected]
  http://clam.iua.upf.edu
_______________________________________________
CLAM mailing list
[email protected]
http://clam.iua.upf.edu

Reply via email to