BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Enhancing Signature-based Collaborative Spam Detection - Jeff Yan\
 , University of Newcastle upon Tyne
DTSTART:20060331T150000Z
DTEND:20060331T160000Z
UID:TALK5482@talks.cam.ac.uk
CONTACT:Saar Drimer
DESCRIPTION:To date\, statistical spam filters are probably the most heavi
 ly studied\, and most widely adopted technology for detecting junk emails.
  However\, among other disadvantages\, they fail to detect spam that canno
 t be predicated by machine learning algorithms on which they are based. Ne
 ither they identify spam that is sent in an image format. In addition\, th
 ese filters need to be regularly trained\, particularly when false positiv
 e occurs. Signature-based collaborative spam detection (SCSD) seems to pro
 vide a promising solution addressing all these problems. What is in partic
 ular attractive is that it can provide a reasonalbe solution to detect unf
 oreseeable new spam\, which intuitively appears to be mission impossible. 
 In this talk\, I will discuss reesarch issues in SCSD\, and report our enh
 ancements to two representative systems\, Razor and DCC. One key problem a
 ddressed by us is that SCSD approaches usually rely on huge databases of e
 mail signatures (i.e.\, checksums)\, demanding lots of resource in signatu
 re lookup as well as signature database storage\, transmission and merging
 . In our enhancements\, signature lookups can be performed in O(1)\, i.e. 
 constant time\, independent of the number of signatures in the database. S
 pace-efficient representation can reduce signature dababase size by a fact
 or of 25.6 or more for Razor-style systems before any data compression alg
 orithm is applied. A simple but efficient algorithm for merging different 
 signature databases is also supported. If time allows\, some ongoing work 
 and open problems will also be discussed.
LOCATION:FW11\, Computer Laboratory\, William Gates Building
END:VEVENT
END:VCALENDAR