Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Next revision Both sides next revision
services:e-mail_service:antispam_feedback [2009/12/16 12:35]
marina
services:e-mail_service:antispam_feedback [2015/08/17 14:07]
giunta
Line 1: Line 1:
 ====== Anti-Spam Feedback ====== ====== Anti-Spam Feedback ======
  
-**To improve the spam filters we need a feedback by the users.**\\ +To improve the spam filters we need a feedback by the users.\\ 
-**False positives** are legitimate e-mail messages that are mistakenly blocked or tagged as spam by the anti-spam filter . \\ +  ​* ​**False positives** are legitimate e-mail messages that are mistakenly blocked or tagged as spam by the anti-spam filter .  
-On the other hand there are e-mails (**//false negatives//**) that pass the anti-spam filters and arrives on the users' inbox, but they are spam.+  ​* ​On the other hand there are e-mails (**false negatives**) that pass the anti-spam filters and arrives on the users' inbox, but they are spam. 
 ---- ----
  
 Our internal anti-spam [[antispam_filters#​barracuda_spam_firewall|Barracuda Spam Firewall]] assigns to every incoming mail **a spam score value**. Our internal anti-spam [[antispam_filters#​barracuda_spam_firewall|Barracuda Spam Firewall]] assigns to every incoming mail **a spam score value**.
-E-mails with values higher than 3.5 are probably SPAM messages and the anti-spam tags their __subject line__  ​with the identifier ​//**[Maybe SPAM]**// +E-mails with values higher than 3.5 are probably SPAM messages and the anti-spam tags their //subject line// 
 + with the identifier **[Maybe SPAM]**
  
-__**Only ​the mails**  ​with very-very-high spam score (**spam score over 9 points) are blocked by the Barracuda Spam Firewall**__. \\ +**Only ​e-mails with very-very-high spam score (over 9 points) are blocked by the Barracuda Spam Firewall**. \\ 
-**The others --//allowed or tagged as probably spam//-- are delivered to the final user. +The others --//allowed or tagged as probably spam//-- are delivered to the final user. 
-**\\+\\
  
  
-  * See here about __[[webmail-horde:​filter:​spam-level| how to manage my e-mail spam filters using HORDE ]]__ +  * See here about [[webmail-horde:​filter:​spam-level| how to manage my e-mail spam filters using HORDE ]]. 
-  * See here about __[[webmail-squirrel:​filter:​spam-level| how to manage my e-mail spam filters using SQUIRRELMAIL ]]__ + 
-  * See here for __[[services:​e-mail_service:​Some-definitions|some useful answers]]__ to questions about spam.+/** inizio commento 
 +See here about [[webmail-squirrel:​filter:​spam-level| how to manage my e-mail spam filters using SQUIRRELMAIL ]]
 +fine commento */ 
 +  * See at the end of this page about some useful answers to questions about spam.
 \\ \\
 ---- ----
Line 28: Line 33:
  
 When you find a:\\ When you find a:\\
-**false positive** --mail tagged as SPAM or "[Maybe-SPAM]"-- which is NOT a real spam \\+  * **false positive** --mail tagged as **SPAM** or **[Maybe-SPAM]** -- **which is NOT a real spam** 
 or\\ or\\
-**a spam** mail without ​"[Maybe-SPAM]" ​in the subject,\\ +  * **a spam** mail** without [Maybe-SPAM]** in the subject, 
-please [color=magenta]**forward it, with its complete header, to [[spam@sissa.it]]**[/color] [color=magenta]specifying[/color] ​if it is a FALSE POSITIVE or a SPAM.+**please ​forward it, with its complete header, to [[spam@sissa.it]]** ​**specifying ​ in the //subject line//​** ​if it is**FALSE POSITIVE** or a **SPAM**.
  
 The operators take the notifications and update the filters. The operators take the notifications and update the filters.
Line 38: Line 43:
 We will take the notifications and update the filters, in order to make the filtering better. We will take the notifications and update the filters, in order to make the filtering better.
 \\ \\
-You find here __[[services:​e-mail_service:​header-example|typical ​header example]]__.+You can find a [[services:​e-mail_service:​header-example|header example ​on the page: typical header]].
 \\ \\
 You find below some instructions to correctly send us the complete header. You find below some instructions to correctly send us the complete header.
  
  
-=== Using PINE to send us the complete e-mail header ====+=== Using ALPINE ​to send us the complete e-mail header ====
  
 You must enter/exit the //header mode// with the key **H**. ​ You must enter/exit the //header mode// with the key **H**. ​
-When you enter the header mode The program shows at the bottom of the screen ​the [Display of full headers is now off.] or [Display of full headers is now on.]+When you enter the header mode, the program shows at the bottom of the screen ​an information message like: 
 +  * //Display of full headers is now off//  ​or 
 +  * //Display of full headers is now on.//
 Now you can forward us your e-mail. ​ Now you can forward us your e-mail. ​
 \\ \\
Line 52: Line 59:
 === Using Thunderbird to send us the complete e-mail header ==== === Using Thunderbird to send us the complete e-mail header ====
  
-Open the message and then select the menu ''​**View --> Headers --> All**''​: you will see on the message window the complete headers of the messages.+Open the message and then select the menu **View** --> ​**Headers** --> ​**All**: you will see on the message window the complete headers of the messages.
 Now you can forward us your e-mail. Now you can forward us your e-mail.
 \\ \\
 +
 +---- 
 +
 +===== Some Useful Answers to SPAM Questions =====
 +
 +**What are false positives?​**\\
 +False positives are legitimate e-mail messages that are mistakenly rejected or filtered by a spam filter.
 +
 +\\
 +
 +**Will I get false positive?​**\\
 +Yes: for any false positives the sender will receive a bounce message
 +indicating the email did not go through. ​
 +Your notification to [[spam@sissa.it]] of the false positive allows us to train the Bayesian filters and
 +further reduce the occurrence of false positives.
 +
 +\\
 +
 +**What is Spam scoring, and the Bayesian Analysis?​**\\
 +  * Spam scoring is the rules based system (about 4000 rules) that looks for characteristics in the email. The Spam scores are used to determine what to do with the email. They do not contribute directly to the future analysis. The rules are updated via the Updates Subscription that we maintain on Barracuda.
 +  * Bayesian analysis, also contributes to the Spam scoring. This is done using your notifications,​ so that we can flag some messages as either SPAM, or NOT SPAM in the message log.
 +
 +\\
 +
 +**How Can I find out if an email that I am missing was blocked by Barracuda?​**\\
 +You find in the Junk folder the e-mails stopped by Barracuda because of spam. You receive an e-mail notification concerning e-mails stopped because of viruses.
 +In case you have more questions you can contact the [[helpdesk@sissa.it|ITCS Helpdesk]] and we will check the message log for you.
 +
 +