Feedback loop for spam prevention

The present invention provides a feedback loop system and method that facilitates classification of elements associated with spam blocking in server-based and / or client-based architectures. The invention employs an automatic learning method used on spam filters, and specifically selects incoming emails randomly, so that examples of both legitimate email and spam / spam are used to generate sets of training data. Users identified as spam fighters ask to rate whether a selection of their incoming emails is either legitimate or spam. A database stores the properties of each record and rating transaction such as user information, message properties and summary of content, as well as query results for each message to generate training data for automatic learning systems. The automatic learning systems make it easier to build improved spam filters that are trained to recognize both legitimate and spam, and to distinguish between them.