Spamihilator sits between your email client (Outlook, Thunderbird etc) and the internet and examines every incoming email in order to identify spam. It employs a Baysian “learning” filter that calculates the probability that a certain message is spam, and that gets better the more you use and ’train’ the program.
[Editor’s note: this review was written by reader Carbonize. Check out his tech blog here].
Spamihilator uses a Bayesian learning filter the same as most spam filters do. It also supports both the blacklisting and whitelisting of email address. Fine nothing new there. Where Spamihilator does excel is that it supports plugins and comes with several filters by default.
- Learning Filter: analyses the words in the email and assigns them a score depending on if the email is spam or not. This is then used to give a score to incoming emails to see if it looks like spam or not.
- Spam Word Filter: You can also add your own words to look out for and say what, by means of percentage, how likely an email containing that word is likely to be spam.
- Link Filter: All links in an email are compared with the black/white list of addresses Spamihilator has stored from past emails. If you say an email is spam then any links in that email get classed as spam links so any further emails containing that link are classed as spam. Like the learning filter there is a scoring system used for links.
- DCC Filter: which connects with the DCC Network to recognize unsolicited bulk mail. The emails are compared with emails received by others and if a lot of other people received the same, or a very similar, email then it gets marked as spam.
- Attachment Filter: marks an email as spam if it has an attachment who’s file type is in the blocked list.
These are just the filters that Spamihilator comes with but you can find more on the site. Personally I just run Spamihilator with the default filters and find it stops about 99% of spam.
All spam emails are blocked and stored in Spamihilators recycle bin which can be accesses from it’s icon in the system tray. From there you can also access the learning area where a list of all received emails is kept along with if they were marked as spam or not. In the learning area you can correct any mistakes such as changing emails marked as spam to not spam and of course email marked as not spam to spam and then click ’learn’.
Spamihilator supports both POP3 and IMAP as well as SSL connections for both. It also has a feature called Cross Protocolling which means if your email client only supports POP you can still use it to retrieve email via IMAP.
CONS: The DCC filter can produce a lot of false positives for things like newsletters, password reminders and sign up emails from popular sites and software manufacturers.
Version Tested: 0.9.9.44
Compatibility: Windows 2000/2003/XP/Vista.
Go to the program page to download the latest version (approx 1.9 megs).
Nice one. Bayesian filter is one of the best approaches because it actually learns from user (instead of dumb “our super-smart technology”).
I mostly use Opera M2 for my mail and it has native Bayesian filter, works excellent.
Actually it’s my experience that most spam gets caught by the DCC and links filters.
I forgot to say in my review that Spamihilator lets you specify what order the filters are used in.
No link exists to the program page!
LMAO you’re right – http://www.spamihilator.com
Oops. Added the link to program page.
I tried spamihilator once.. but I always comeback to Spambayes. I think it has a little bit issue with Outlook 2007.
Because Spamihilator sits between your email client and the internet it will work with any POP3 or IMAP email client. It does not interact with the client in anyway and so Spamihilator cannot affect the client in any way.
I think its a good product. But need to test on it.