Spam Terrier detailed features:

Extensive mail control

Personalized filter training

Spam Terrier’s Bayesian filter, the core processing engine, can be trained to improve spam recognition efficiency with each incoming email.Because spam is different for different people, you get to define what should be considered spam for your purposes.

By training Spam Terrier on what constituted spam in the past, you can demonstrate to the program the characteristics (or definite such as "cheap Viagra") of spam that can be applied to your future correspondence. The filter then learns certain behavior/appearance patterns and applies this knowledge to categorize future mail.

The Bayesian filter is based on probability analysis assessthat assesses the odds of new messages being spam that have similar characteristics to those of past emails designated as spam. The more thoroughly you train Spam Terrier, the more accurate and tailored to your requirements the results will be.

More efficient database compilation and management

As a result of a number of internal optimizations to the database structure, the spam filter is now easier to train, update, manage and maintain. All these measures deliver better overall spam detection efficiency. Here’s what’s been improved:

• Time-based database separation
  The spam knowledge base is now broken out into segments according to the date of the last user input. More recent input is given greater weight, which helps the program to better adapt to the dynamic nature of spam.
• Faster and more granular processing
  When Spam Terrier debuted, it definition database to process incoming mail. Now the database has been broken out into multiple pieces to speed up processing, and provide more efficient management and updating. As a result, Spam Terrier is now faster, more responsive and uses less CPU time and system memory.
• Database size balancing
  Because of the possibility of imbalance in the spam/not-spam email messages used during the training phase, there was potential for one category (i.e. definite spam) lagging behind the other (definite non-spam) in the amount of available classification data. To deal with this situation, Spam Terrier now discards messages beyond the number of emails constituting the minority group, eliminating the so-called database "tilting" effect.
• Accelerated learning from the interface
  Now Spam Terrier accommodates user input in real time, making spam detection even more accurate

Adjustable sensitivity thresholds

Each incoming message is assigned a "spam score" by Spam Terrier, which classifies the message as not spam/probable spam/definite spam. You can change the way these calculations are made by modifying the values the scores fall into to suit your level of tolerance.

Definable whitelists and blacklists

Spam Terrier doesn’t limit its filters to the use of senders’ white- or blacklists (respectively, the list of allowed/blocked senders) when it processes emails. Users can also create special filters based on a number of variables: for example, message body text, header content, or email subject. Email addresses in users’ address books are automatically added to the list of legitimate senders.

No lost messages

Suspected spam messages are put into a special folder that you can check at your convenience and restore any misclassified messages. You can also expand the Spam Terrier Bayesian database by clicking "mark as not spam" on any valid message in the Spam folder.

View message status

By clicking on any message in any folder, users can see how likely that message will be flagged as spam by reviewing its cumulative spam score. Users also thus get valuable insight into how the ranking works in practice.

Continuity of protection

Automated spam check

Spam Terrier’s real-time self-learning algorithm ensures every incoming message is automatically analyzed against the program’s indicators to determine the legitimacy of the message. Suspicious emails are then quarantined for later review.

As Spam Terrier constantly scans new mail in search of fresh spam samples on which to base its junk mail database updates, the accuracy rate increases over time and minimizes the number of false positives while maximizing the spam identification rate.

On-demand scan

Spam Terrier can check existing email folders for the presence of spam. This on-demand spam scan removes junk emails in advance from any specified folders, and ensures that mailboxes start out as clean as possible.

Ease of use

64-bit email clients support

Few commercial antispam solutions support the 64-bit software environment, let alone free products. Agnitum Spam Terrier was developed specifically for 64-bit Windows platforms, enabling you to benefit securely and fully enjoy from the performance and expandability benefits of 64-bit computing.

Intuitive installation

The process of installation is so easy that even a beginner will have no problem getting started with Spam Terrier.

Seamless integration with email clients

Spam Terrier is integrated directly into most popular email clients, including support for The Bat! and Windows Mail applications. It adds a small toolbar to the menu bar of your favorite email program. Controls are intuitive and no special knowledge is required.

Support for multiple email accounts

Spam Terrier supports multiple email accounts when it checks spam.

Multiple-language interface

Spam Terrier is available in English, German, Spanish, French, and Russian.