Home
| Calendar
| Mail Lists
| List Archives
| Desktop SIG
| Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU |
On Tue, Feb 11, 2003 at 11:48:46AM -0500, David Kramer wrote: > > > Does anybody know if someody has written a logic filter for email > > > hoaxes? Would be interesting as they all seem to sound the same to me. > > > > Unfortunately, it would require the ability to parse English, a > > strong-AI condition. > > Actually, yes. A friend of mine wrote one. > http://crm114.sourceforge.net/ Sorry. As Crash will be the first to tell you, the CRM-114 Discriminator doesn't understand English, and can't parse English. All it does is match word-probability against the existing corpus-profile. In other words, there are several approaches which will work, but none of them are "logic filter"s that analyze the logic of the statements made. -dsr- (who did some very minor debugging on it before it was released, or even named.)
BLU is a member of BostonUserGroups | |
We also thank MIT for the use of their facilities. |