Spam Auto-Learning

[m1byo]

Would it be possible to include an option to allow for the Baysean Auto-Learn feature to only learn from emails marked as Read in addition to the current features

At the moment, I mark all of the emails in my Junk folder as read just to signify that I have had a quick glance through the emails to ensure there are no emails which have been mistakenly identified, this feature would allow for the correctly identified spam to be added to the database a lot sooner.

thanks

Ian

[m1byo]

Also, in this sort of area! would it be possible to include a way of marking any emails which were mistakenly identified as spam as non-spam immediately so you dont have to wait 3 days (or longer depending on the auto-learn settings) before the filter will resolve its mistakes!

Thanks

Ian

[rob]

Only learning from read mail is quite a wise idea in that it would ensure only mails that users have checked will be used (reducing the chance of poisoning the filter). I have made a note of your suggestion and we will consider it for a future update. As to your second suggestion, during design we did consider an instant processing method but of course the delay ensures that users get plenty of time to organise their mail. For mistake corrections, the original date the mail arrived will remain as is and so when you move the mail, the next auto-learn phase (2.30am by default) will cause it to be relearnt.