ASSP is currently re-inventing the wheel by building, optimizing, and updating a bayesian filter.
Would it be worthwhile to explore replacing that engine with DSPAM? DSPAM is a very advanced bayesian filter that has (I believe) many more features and is potentially a lot smarter than the bayesian filter in ASSP. It also learns via incoming and outgoing mail, and can be trained via an email interface, just like the current filter. A few of the methods it uses are outlined on the website at http://dspam.nuclearelephant.com/ "Some of the approaches deployed in DSPAM include Concept Identification, Neural Networking, Message Inoculation , advanced de-obfuscation techniques, and a new noise reduction algorithm called Bayesian Noise Reduction. DSPAM also supports many different mathematical paradigms including Bayes, Chi-Square, Geometric, and Markovian Discrimination." The core engine is available for dropping into products like ASSP at no charge "For developers, the DSPAM core engine (libdspam) can be easily incorporated directly into applications for drop-in filtering (GPL applies; commercial licenses are also available)." The README for the latest version (3.8.0) isn't online for some reason, so I've uploaded the entire package here: http://www.urigubu.com/dspam-3.8.0/ See the README and /src/example.c for libdspam integration details What does everyone here think about this? David ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Assp-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-user
