Personalized Collaborative E-mail Filtering Model Based on P2P Network
Guangshun Li · Computer Technology and Development · 2011
E-mail is filtered with user's feedback information about false positives and false negatives in content-based collaborative filters.While this improves the ability of the filters to track concept drift in spam over time,such approaches make implicit assumptions in centralized spam filtering.A personalized collaborative spam filtering model based on P2P network was presented to solve the problems in centralized collaborative spam filters and the design and implementation of signature-based system and the architecture were described.A prototype system was implemented to analyze the data collected from the users employing the system as their spam filtering tool.The evaluation of the results indicates that the personalized collaborative spam filtering is accurate and efficient.