Neat stuff for the day

Paul Graham: Better Bayesian Filtering

If people had been onto Bayesian filtering four years ago, why wasn't everyone using it? When I read the papers I found out why. Pantel and Lin's filter was the more effective of the two, but it only caught 92% of spam, with 1.16% false positives.

When I tried writing a Bayesian spam filter, it caught 99.5% of spam with less than .03% false positives.

So why did we get such different numbers? I haven't tried to reproduce Pantel and Lin's results, but from reading the paper I see five things that probably account for the difference.

John Udell: Shipping the Prototype "Let's promote scripting languages to the status they deserve"

Check out this neat interview with Dennis Richie.

Permalink | Comments? | id 3421

21:13 (utc-5)
edited: 22:37 (utc-5)

 

Firefly Theme

It's Friday, and Firefly would have been on in about a half hour... I've been listening to the Firefly theme today and getting depressed... I miss Firefly. I hope it comes out on DVD.

Permalink | Comments? | Tags: TV/Movies | id 3420

19:30 (utc-5)

 

Chandler alpha due by April

Turns out a Chandler alpha may be out by April.

Kapor, who spoke to about seventy Stanford University students as part of the university's computer systems colloquium, said the project's goal was to do for application software what Linux had done for operating systems: begin life as essentially a group project, then later challenge proprietary systems and win.

"We think this is a good time to bring to use the OS model to bring to the world of applications software of uncompromising quality," Kapor said.

The group has begun architectural coding, and hopes to have an early alpha release available in April, with bug reporting capabilities and perhaps some networking and security features, Kapor said. In a November blog entry, Kapor predicted that "optimistically, we could have a 1.0 by the end of 2003. Pessimistically, it will be 2004."

Chandler contains both a client and server on the same machine, but said a hosted server version could be added later to challenge Exchange on its home turf, Fortune 500 companies with thousands of Exchange accounts. The software uses several open-source packages, such as Jabber, wxWindows, and Python, as a foundation.

Permalink | Comments? | id 3419

11:46 (utc-5)