Title :
Collaborative Spam Filtering Using E-Mail Networks
Author :
Kong, Joseph S. ; Rezaei, Behnam A. ; Sarshar, Nima ; Roychowdhury, Vwani P. ; Boykin, P. Oscar
Author_Institution :
Dept. of Electr. Eng., California Univ., Los Angeles, CA
Abstract :
Widespread Internet use has led to the emergence of several kinds of large-scale social networks in cyberspace, most notably e-mail networks. Unfortunately, the tremendous convenience that e-mail affords comes at a high price: Although the network´s underlying connectivity is hidden, making it almost impossible to build a comprehensive directory of every individual´s contact lists, e-mail addresses themselves can be easily obtained from publicly available documents. Spammers have exploited this vulnerability to inundate users with unsolicited bulk e-mail. A distributed spam-filtering system that leverages e-mail networks´ topological properties is more efficient and scalable than client-server-based solutions. Large-scale simulations of a prototype system reveal that this approach achieves a near-perfect spam-detection rate while minimizing bandwidth cost
Keywords :
Internet; data privacy; information filtering; information filters; unsolicited e-mail; Internet; bandwidth cost minimization; collaborative spam filtering; e-mail networks; large-scale simulations; large-scale social networks; near-perfect spam-detection rate; prototype system; Bandwidth; Collaboration; Electronic mail; IP networks; Information filtering; Information filters; Large-scale systems; Social network services; Unsolicited electronic mail; Virtual prototyping; e-mail networks; spam;
DOI :
10.1109/MC.2006.257