Contents  SpamSieve Manual  Technical Support

6.3.1   Allow duplicates in corpus

If you allow duplicate messages in the corpus, training SpamSieve with the same message twice will increase the corpus counts for the words in that message. If you do not allow duplicate messages, the second and subsequent trainings with that message will have no effect. By default, duplicate messages are not allowed in the corpus. This is nice because it means that you do not have to remember which messages you have already trained SpamSieve with; accidentally training with the same message more than once will not skew the data that you are providing to SpamSieve. If you wish to intentionally skew the data, you can check one or both boxes to allow duplicates.

     Contents  SpamSieve Manual  Technical Support