This embeddable tool works with Outlook and helps detect spam. While it can detect known methods of spamming, it can be further trained based on the choices made by a user to mark a message received as spam or not.
Pros: The application moves ambiguous messages to a quarantine folder besides the usual junk mail/bulk mail folder. Based on the user choices of the quarantined messages, the filter gets further training inputs for determining what’s spam and what’s not. While this builds up the knowledge base the application can actually start work right away as it comes with a knowledge base built in and does not require any configuration of Outlook. Available wizard takes care of configuring the tool. The knowledge base includes known spam types and they can be detected right away after the tool is installed.
The basic filtering is based on Bays statistics as most these spam filters are. Publishers claim that many more parameters are taken into account when determining what is spam. While that is difficult to verify, the effect should be visible in filtering the results. The algorithm is claimed to be tuned to reducing the probability of false positives.
The tool adds a tab on the tools/options menu. The statistical date of how many proper mails, how many spam according current knowledge base were received, are indicated. False positives and false negatives statistics are also made available, as is % messages whose status could not be determined. These statistics indicates how well the application has been trained. The statistics can be reset and a new cycle started from the last reset.
Cons: Nothing adverse noticed during evaluation.
Overall: Real evaluation of this kind of tools is by their performance. That takes quite a bit of time to build up the knowledge base and train the application well. Based on the way the application works and how easy it is to use and whether effective setting can be made are the criteria for evaluation here. Based on these the overall rating is 4 stars for the tool.