signalnine dot net

The Press Release.
The Demo.

Please test it out and let me know what you think.

Here.

Is it *really* necessary to implement a text classifier as a Turing-complete programming language? With declensional syntax? WHY GOD WHY

Just posted here [mp3 download 11MB]. My interview starts at about 7:40.

We’ve been invited to ROFLcon. We are so there.

So, Fox Business News wants to do a live interview with me on “Money For Breakfast” sometime this week.

What the hell should I wear?

The Fortune magazine article just posted. BBC interview is scheduled for Thursday morning.

First, Fortune Magazine, now the BBC wants to interview me regarding StupidFilter. More information as soon as I’ve got the interview set up.

Crazy.

Done:

• Now have a pretty huge MySQL database full of stupid comments, mostly from Youtube225441 rows in set (1.21 sec)
• Implemented a web-based comment ranking system to seed the stupidity corpus
• Tracking moderator history, comment rankings, dropping spam and unusable comments
• Recruited the celebrated Paul Starr, incomparable Jess, esteemed Sparky, lovely wife Lexi, and [your name here, apply within] to serve as Moderators Extraordinaire

To-Do:

• Hit a minimum of 2,000 usable comments ranked
• Train Bayesian classifier on seed corpus
• Fine-tune filter, begin self-training database using seed data
• Test filter on real world data, tune filter, rinse, repeat
• Implement a web front end for the Bayesian classifier for demonstration purposes
  Edit: Want some random stupid? http://stupidfilter.org/random.php