Q & A Home
Version 3 Architecture
What happens after I submit a false positive?
Please refer to our FP handling process for an overview of the complete process.
We take false positives very seriously - one of the key values of our system is the extremely low rate of false positives. However, every system has different needs so spam for one subscriber may be a false positive for others. Each Message Sniffer rulebase starts from a single common database which we update continuously and then evolves for the specific needs of the target system. Typically a new system requires a week or two of adjustments for best performance, however, many systems find the standard rulebase to be sufficient without any additional tuning.
If we find that a rule in the core database is causing a significant number of false positives or otherwise determine that it should be refined, then we will take the appropriate action. However, if the false positive report is largely isolated to your system then we may leave the rule in place and create an appropriate bypass in your specific rule base.
In some cases, it may still be preferable to make adjustments in your local system. For example, in Declude Pro you can adjust the weights of different tests, adjust processing rules for single domains or users, and use the WHITELIST ANYWHERE feature to override all tests for specific content - such as a newsletter name or some consistent component of the message header. Your specific software may have similar capabilities that you wish to exploit in customizing your anti-spam solution. You may even wish to use multiple Message Sniffer licenses with different rule bases for specific purposes.
IMPORTANT: Ebay, Yahoo groups, and other lists frequently include advertisements that may trigger matches in sniffer's rulebase. While we are creating standard white-rules to mitigate the effects of this, it is important that you establish some appropriate white-rules for your system when you first begin to use Message Sniffer so that you can tune it to your specific needs and avoid these and other lists that may be falsely identified due to their advertising content.
Since the release of Version 2 we have begun implementing our own "standard" white-rules to mitigate this problem. None the less, some specialized tuning will always be necessary because each individual system has it's own needs and definitions for spam / not-spam and some trade-offs cannot easily be made. For example, we have attempted to create a white-rule for hotmail accounts, unfortunately this produced a significant increase in uncaptured spam so the rule was removed. Depending upon your needs you may wish to make this trade-off in favor of white-listing hotmail accounts.
Our support staff will help you customize your rulebase to fit your needs.