If a filter correctly identifies a type of spam and auto-deletes it, does MW also evaluate those emails for learning purposes?
AJ
how filters affect learning
- Sidewinder
- Weary Womble
Post
Re: how filters affect learning
Not sure with regard to the auto delete but Learning is normally controlled based on the "Total" evaluation numerical result. There are two ranges regarding the actual number that affect the decision:
The Total number is also influenced by the Filter slider:
The mathematical summation of the numerical values of each of the MWP evaluation tools as shown in the Sidebar determine the Total scoring for the message and whether or not it will be learned.I am not a Firetrust employee. Just a MW User & Volunteer BETA Tester.
Remember "FREEDOM IS NEVER FREE" U.S.N.
DT W7 64 HP SP1 16GB Ram - LT W7 32 HP SP1 4GB Ram - iPad4 64 GB Ram WiFi/Cellular IOS 9.3 Beta 3
Remember "FREEDOM IS NEVER FREE" U.S.N.
DT W7 64 HP SP1 16GB Ram - LT W7 32 HP SP1 4GB Ram - iPad4 64 GB Ram WiFi/Cellular IOS 9.3 Beta 3
- Sidewinder
- Weary Womble
Post
Re: how filters affect learning
See this topic regarding the Learning Range Values for the Total scoring:
http://forum.firetrust.com/viewtopic.ph ... 401#p23815
http://forum.firetrust.com/viewtopic.ph ... 401#p23815
I am not a Firetrust employee. Just a MW User & Volunteer BETA Tester.
Remember "FREEDOM IS NEVER FREE" U.S.N.
DT W7 64 HP SP1 16GB Ram - LT W7 32 HP SP1 4GB Ram - iPad4 64 GB Ram WiFi/Cellular IOS 9.3 Beta 3
Remember "FREEDOM IS NEVER FREE" U.S.N.
DT W7 64 HP SP1 16GB Ram - LT W7 32 HP SP1 4GB Ram - iPad4 64 GB Ram WiFi/Cellular IOS 9.3 Beta 3
- allangjohnson
- Travelling Tuatara
Post
Re: how filters affect learning
Thanks, though I have to admit the technicalities of this are a bit beyond me - my main concern is whether having filters in place interferes with the MW learning function.
AJ
AJ
- stan_qaz
- Omniscient Kiwi
- Location: Gilbert, Arizona
Post
Re: how filters affect learning
Nope, I have a huge pile of filters and for me they help learning since they weight the spam correctly in many cases saving me having to click the learning buttons to correct them. If I spit 500 test messages into MW with the learning tool reset to a blank state the first wash (using just the filter results and no manual classification) populates the tables with enough info that it is back to a very respectable level of accuracy. If I go through and deal with the ones not filtered or filtered to neutral the accuracy goes even higher.
I don't know about auto-delete's effect and likely will never risk using it to find out.
I don't know about auto-delete's effect and likely will never risk using it to find out.
- QuietOne
- Omniscient Kiwi
- Location: Texas, USA
Post
Re: how filters affect learning
The short answer to your question is "No".
The filters you create will only help you organize and designate any email you receive as "Good", "Spam", or "Neutral" (i.e. Undecided). The Learning goes on behind the scenes and will take into account what filters you write and the various messages caught by them but is otherwise separate and will run just as well w/ or w/o filters.
The filters you create will only help you organize and designate any email you receive as "Good", "Spam", or "Neutral" (i.e. Undecided). The Learning goes on behind the scenes and will take into account what filters you write and the various messages caught by them but is otherwise separate and will run just as well w/ or w/o filters.
Stealth is the Best Weapon.
15" MacBook Pro, OS X 10.14.6, 16GB RAM, Parallels install went kaput so no Windows version, Mozilla Thunderbird. All T'bird accnts: POP3. iPhone Mail Accounts IMAP. Unsubscrbd DNSBL.
And I'm: Just Another βeta-Tester.
15" MacBook Pro, OS X 10.14.6, 16GB RAM, Parallels install went kaput so no Windows version, Mozilla Thunderbird. All T'bird accnts: POP3. iPhone Mail Accounts IMAP. Unsubscrbd DNSBL.
And I'm: Just Another βeta-Tester.
- allangjohnson
- Travelling Tuatara