I write to inform you of new enhancements we will be making to the campus anti-spam program in June and July.
Some of you have expressed concerns regarding the quantity and nature of unsolicited spam getting through to your email in-boxes. For the last two years, Information and Educational Technology (IET) has been running with some success an anti-spam filtering system on the campus email servers. Additional opt-in services are also provided for those interested in configuring their email programs to catch more of the spam sent to their accounts. Despite this program, spam continues to rise at UC Davis as well as throughout the Internet community.
To address this on-going problem, we have been investigating additional anti-spam measures that we are now preparing to roll out. In brief, with these measures, we propose to attack the problem of spam on several fronts:
- A real-time "blocking" list will be established in June that will capture the off-campus IP addresses from which high levels of spam are sent. Email messages originating from those IP addresses will then be automatically rejected by the campus email servers.
- In July, we will refine our anti-spam approach with better bulk email detection capabilities (using the Distributed Checksum Clearinghousesystem) and with a type of filtering (called Bayesian filtering) that learns from spam and from legitimate mail, resulting in more robust, adapting and efficient anti-spam filtering that typically returns few false positives.
- Also in July, we will roll out a mechanism to temporarily quarantine high-scoring messages. With this system, messages scoring 15 or higher will be sent to special folders automatically created for all campus email users. Quarantined messages delivered to these folders will be deleted after 28 days. In the future, we will analyze the feasibility of dele ting these messages to avoid virus inspection, delivery and storage of undesirable messages.
Clearly, defining exactly what constitutes spam for a group as large as the UC Davis community, and fighting spam efficiently, remains a difficult challenge. As we all know, what one considers spam another may consider legitimate email. Additionally, over time, characteristics of spam may change as spammers discover new ways to evade filtering programs, and the efficiency of spam detection and filtering programs may fluctuate accordingly. With our proposed enhancements, however, we will expand, strengthen, and build greater flexibility into our anti-spam program, thereby minimizing disruptions to campus email users.
For additional information about campus spam filtering services, visit http://security.ucdavis.edu/spam.cfm.
Peter Yellowlees, MBBS, MD
Interim Vice Provost
Information and Educational Technology