Anti Spam - SpamUnit

Spam filtering software is available in numerous forms. It can be part of an email client or a plug-in for an existing application. Most share common features, with rulesets, block lists and Bayesian spam filtering all common features of spam filtering software.

First of all, spam rules (or “filters”) are commonly used and are heavily customisable. Most antispam software lets you filter out email manually - any number of features of the email can be used in the filtering process, from the originating address, to the subject line, to certain words included in the message. Similarly, the technical information in the email header - including IP and routing information - is also commonly used when filtering out unwanted email. Some spam filters also allow geographic filtering, with certain countries being excluded from your inbox automatically. Spam lists or block lists are more commonly used by email server administrators, but can also be employed in spam filtering software, particularly in conjunction with the geographic and header filters.

In addition to blacklisting certain addresses and other identifying information, users can also whitelist trusted senders. Occasionally, a welcome, innocent message can be rerouted to the Spam folder by antispam software. (That’s why it’s important to keep an eye on your spam folder, just to make sure your spam protection software is operating correctly.) Of course, anyone can move an email from the junk mail folder to their inbox, but most antispam software allows you to then automatically whitelist all messages - past and future - from that particular address so that the same thing doesn’t happen again. Similarly, you can pre-empt such problems with contact lists and friends lists supported by some email clients, webmail clients and almost all antispam programs.

As well as recognising spam patterns, some antispam software has dedicated phishing detection algorithms. Phishing often takes a different pattern to traditional spam emails and often requires a different method of detection. Phishing mails can look like legitimate emails from corporations or even like traditional letters, so dealing with them separately is a must - pattern recognition and a list of websites used to facilitate phishing fraud are commonly used together to sort these phishing attempts out from legitimate messages. Similarly, image spam is a growing problem and spam software has had to deal with it. Images from non-trusted senders can be blocked until the user enables them, while image-laden emails are quite often immediately recognised as spam and filed accordingly.

It’s the automated spam detection that often makes or breaks these packages. Many antispam programs now use Bayesian spam detection techniques, which first came into common use around 2002 to provide a new alternative to detecting junk email content. It detects and stores commonly-received spam terms unique to the individual user to help detect what is truly unwelcome mail. Because it works on an individual’s own mailbox, rather than a centralised server, it is often more reliable than a catch-all solution. What’s more, rather than simply filtering out individual words or email subjects, it will only filter them out in conjunction with other words, phrases and patterns that make up a spam email. Filtering out the word “Viagra” could (very) occasionally result in missing a legitimate message, but using Bayesian filtering to remove Viagra in conjunction with certain other tell-tale signs of spam is a much more flexible, reliable and user-friendly solution.

Some spam filtering software also makes inroads into the spam problem by trying to remove your address from spammers’ contact lists. Creating an automated “address not found” error message is a common way of discouraging some spammers, as they may then think the address is inactive and remove it from their systems. Similarly, some programs automate the abuse reporting procedures, allowing you to quickly and easily contact ISPs and hosts to alert them of the spam problem taking place from their systems. Again, software packages vary in how much they can do here, but most offer at least some of these features.

There is so much antispam software on the market that looking at the key features that best suit your needs is the only way to go about finding one that suits you. Research is key if you’re serious about stopping spam.