Using trust to block the spread of spam
Electronic mail, or e-mail, has had profound implications for communication for both business and pleasure. Yet email is not without its problems. The spread of spam, unsolicited email, is on the rise. An Information Society Technologies Programme project entitled SECURE advanced the concept of trust in computing, namely exploiting trust to improve security. The SECURE workplan targeted the development of a spam filter. This task was headed by security experts at the University of Geneva. Each new incoming email received by the client's mail server is checked according to the trust model implemented in the SECURE framework. Email emanating from valid sources is sent directly to the user's inbox while the remaining messages are tagged as spam and sent to a special folder. Should valid email be incorrectly classified as spam, or vice versa, the filter monitors the user's actions (e.g. moving messages between folders) and adjusts the trust policy accordingly. Several thousand lines of Java programming enable the spam filter to collaborate with the popular Simple Mail Transfer Protocol (SMTP) on the server side and the Internet Message Access Protocol (IMAP) on the client side. The filter can also be used with version 3 of the Post Office Protocol (POP3), but functionality is reduced. At any rate, IMAP is nowadays available on all major email client applications (e.g. Microsoft Outlook, Mozilla, etc.). The new spam filter will bolster existing spam filters that block email with suspicious content, freeing email users from the dreaded task of deleting endless spam. The University of Geneva plans further testing and development before pursuing commercial exploitation of the new filter.
