Rules

---

The lists of black and white rules allow recognizing spam and non-spam by specific combinations of letters in body and headers of the messages. The rules are the optional supplement for very effective trainable algorithm. But the accurately assembled lists of rules may improve the classification quality in cases when the algorithm is unable to select a certain class for a message with high probability.

The rule consists of one ore more conditions. The rule works when all conditions are satisfied. The rule also works when at least one of the Strong conditions is true.

White rules recognize non-spam and they are checked first. The message is classified as non-spam with rating 0 if at least one white rule worked.

Black rules recognize spam. The message is classified as spam with rating 100 if at least one black rule worked.

You can exclude the rules from consideration by unchecking them in a list. The same feature is available for conditions.