Spam Filtering System Based on Collaborative Filtering
Zou Heng-ming · Jisuanji gongcheng · 2008
This paper proposes a collaborative-based spam filtering scheme to overcome the quality evaluation problem existed in current spam filtering techniques.The scheme can be used to solve the problems with current techniques in classifying Email attachments.Because spam senders often resort to mass mailing,the system improves filtering accuracy by considering the factors that include user actions and reading speed as well as feedbacks from others on the same mailing list.The collaborative-based spam filtering is implemented as part of a hybrid spam filtering system.Experimental results show that the scheme significantly reduces the recall rate.