Junk mail filtering...once again

[kevinp]

Appreciate the suggestion, but I can't add all the contacts from all my clients companies to my address book.

From what I am hearing, on the help that people have given me is, the Bayesian filters in Barca/Poco are not that great, and it's better to use custom filters to only accept email from people in your address book.

Having the filter is good, but not great, the bayesian filters should do a better job. As a test, I think I am going to disable the internal bayesian filters and run a 3rd party filter again, then based on it's scores filter that message. If the internal Bayesian filter is just as good as 3rd parties, I should receive the same results.

I've talked to colleagues that use Thunderbird, and the built in spam filter in thunderbird has pretty much eliminated their spam from their inbox, so it is possible.

I think the developers may need to spend a little bit more time tweaking their bayesian/spam filters. If some people are getting great results, and others are getting poor resutls, something is definitely wrong.

BTW, I have a 71% accuracy on the junk filters at home (Pocomail), and that is VERY poor. This AM when I loaded Poco, 90% of the spam received ended up in my inbox, only filtered 1 email to junk. Something is not working correctly.


Any insight from the Poco Systems developers?

--Kevin

[Michael]

Appreciate the suggestion, but I can't add all the contacts from all my clients companies to my address book.

--Kevin

Kevin:

Using the suggestions here you do not have to add every contact from all your client companies to your address book. Simply add the domains to the accepted sender list. To do this right click on the address of a sender from the company n the message index pane and select "Junk Mail Filtering | Allow Domain". You can also edit the accepted senders file using a text editor, it is the "exceptsenders.txt" file in your "Mail" directory.

[Guest]

I hate to keep flooding this forum with messages, but I am just trying to better understand how all this works.

messages,... Amen to that brother!

~ Link removed ... Eric ~

[Mitch Wagner]

The previous message in this thread is spam ... which is ironic, if you think about it.

[dribnus]

I would have to agree, very ironic and not good There's a couple of post with that link. It's almost like someone is playing with an automated spam program.

[Mitch Wagner]

Not playing ... using. The links appear to be the kind of thing you often see in blog comments spam. The goal of that kind of spam is to raise a page's Google rank.

[Guest]

I have quite good accuracy on my end but then I am running the non-Bayesian filters as well as several processes I've built myself.

"quite" is quite unique, I never heard about something like that before.

~ Link removed ... Eric ~