Best Practices Guide for reducing spam with the Barracuda

This guide outlines the best way to reduce the amount of spam you receive in your inbox by using the Barracuda Spam Firewall.

The First Step

Before customizing your spam setting on the Barracuda, you should setup a rule in Outlook to move all the [RHSPAM] message to the Junk E-mail folder.

  • From the main window of Outlook, click Tools -> Rules and Alerts
  • [Rules and Alerts] Click New Rule...
  • [Rule Wizard] Click Start from a blank rule, choose Check messages when they arrive, and click Next
  • Place a checkmark next to with specific words in the subject then, in the Step 2 area, click specific words
  • [Search Text] Type in [RHSPAM] and click Add then OK
  • Click Next
  • Place a checkmark next to move it to the specified folder then, in the Step 2 area, click specified
  • The Rules and Alerts folder selection window should have appeared
  • Rules and Alerts Click Junk E-mail under the Mailbox - Your Full Name section, then click OK
  • Click Finish then click OK
  • As future messages arrive with the [RHSPAM] in the subject, they will be moved to the Junk E-mail folder

Getting Started

The Barracuda has several ways of determining if an email is spam or not. It uses a combination of real-time blacklists, checks for valid recipients, determines message intent, tests with Bayesian learning filters, and checks personal black and white lists. Some of these filters are system wide and cannot be configured by each user. The Bayesian filter and the personal black and white lists are customizable per user. The most powerful of the customizable filters is the Bayesian learning filter.

The Bayesian Learning Filter

The Bayesian learning filter checks incoming emails based on previous emails you have classified as spam or good emails. After training the filter on at least 200 good and 200 spam email, the filter will be able to reasonably classify most incoming emails.

There are 2 ways to train the Bayesian learning filter. The first is to log in to your Quarantine Inbox (https://rhspam.rose-hulman.edu) and select which held messages should be classified as spam and which should be classified as not spam. This method will allow you to train on several spam messages per day, but very few good emails should ever be in your Quarantine Inbox. This method works best as a way to refine the Bayesian filter.

The second way to train the filter is to use the Outlook Plugin. This plugin will allow you to train the Bayesian filter on messages you already have in Outlook. This is allow you to train the needed 200 good and 200 spam emails more quickly by using emails that have already reached your inbox.

With the needed 200 good and 200 spam email taught to the learning filter, you now have a good baseline provided to the Bayesian learning filter to classify incoming emails.

Incorrectly classified emails (Step 1)

Sometimes an email will be incorrectly classified by the Bayesian filter. When this happens, a good email will typically be held in your Quarantine Inbox (https://rhspam.rose-hulman.edu) or a spam message will get to your inbox.

To correct the misclassification of a good email, login to your Quarantine Inbox, place a checkmark next to the incorrectly classified message, and click the Classify as Not Spam button. You may have to do this for a few similar emails before the Bayesian learning filter will correctly classify them.

By continuing to teach the Bayesian filter what emails are good and which are spam, you should see the best overall results for reducing the amount of spam that arrives in your inbox.

Incorrectly classified emails (Step 2)

(When to use the black and white lists)

If you have tried teaching the Bayesian filter several times about similar messages, but they still are not classified correctly, you may have to use the personal black and white lists.

If the problem is good email marked as spam, you will want to add the sender's email address to your whitelist. When you add an address or domain to your whitelist, all future emails from that sender will be sent to your regular inbox. There are a few ways this can be done.

Method 1) When you receive your Spam Quarantine Summary daily email, there are links next to each email in your Quarantine Inbox. If you click on Whitelist, it will add the sender to your personal whitelist.

Method 2) You can login to your Quarantine Inbox (https://rhspam.rose-hulman.edu/) and either
  a) Click Whilelist next to the message
  b) Click the Preferences tab, enter the email address or domain to the whitelist, and click Add

If spam from the same email address or domain is still arriving in your inbox and you have tried teaching the Bayesian filter several times, then you can add it to your blacklist. Login to your Quarantine Inbox (https://rhspam.rose-hulman.edu/), click on the Preferences tab, enter the problematic email address or domain to the Blocked Email Addresses and Domains (Blacklist) section, and click Add. All future email from that address or domain will be block, so use this feature with caution.