Joho the Blog
|
|
|
May 31, 2004
I've been happily using PopFile to filter my spam in a Bayesian way. But it's inconvenient to train it because it's not integrated into my mail client (Outlook, ulp). I tried Outclass, but it was over-featured and under-explained. So, I'm trying out SpamBayes, another free Bayesian filter with seemingly good integration into Outlook. It doesn't filter into all the buckets that PopFile does, but it's easy to batch-train it against your existing junk and inbox folders. I was motivated to try new filters by the elegant way Thunderbird integrates spam filtering into the client...much as the Mac client does. I'm using Thunderbird on the road, and find much to like about it, not least that it's not Outlook. Posted
by D. Weinberger at May 31, 2004 12:37 PM
TrackBackListed below are links to weblogs that reference Trying SpamBayes:
» Dealing with email spam from Knowledge Jolt with Jack Tracked on June 18, 2004 11:19 PM |
Comments
If you are using Outlook, why don't you give NEO a try?
It sits on top of Outlook and organizes mail much better then Outlook.
http://www.emailorganizer.com/
I am not affiliated etc. I am just a very happy user.
Posted by: Hanan Cohen | May 31, 2004 03:41 PM
Hanan, I did try NEO. I found that I wasn't using it, and that it added too much cruft to the system for my taste. I don't like these low-level add-ins unless I feel pretty confident I know what they're doing. This isn't a criticism of NEO, just an expression of my own neurotic need to be in illusory control of my software.
Posted by: David Weinberger | June 1, 2004 10:09 AM
Good luck with SpamBayes, it works really well for me. After 3 months training I find very few false positives and it siphons off a great swathe of drivel from my inbox. It does seem to slow Outlook down a bit on loading but that's a small price to pay for having Spam expunged from my inbox.
Posted by: Johnnie Moore | June 1, 2004 02:08 PM