期刊名称：International Journal of Computer Science, Engineering and Applications (IJCSEA)
印刷版ISSN：2231-0088
电子版ISSN：2230-9616
出版年度：2014
卷号：4
期号：1
DOI：10.5121/ijcsea.2014.4103
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：Nowadays, an image spam is an unsolved problem because of two reasons. One is due to the diversity ofspamming tricks. The other reason is due to the evolving nature of image spam. As new spam constantlyemerging, filters’ effectiveness drops over time. In this paper, we present an effective anti-spam approachto solve the two problems. First, a novel clustering filter is proposed. By exploring the density-basedclustering algorithm, the proposed filter is robust to spamming tricks. Then, we present a hierarchicalframework by combining the clustering filter with other machine learning based classifiers to furtherimprove the filtering capacity. Moreover, incremental learning mechanism is integrated to ensure theproposed framework be capable of adjusting itself to overcome new image spamming tricks. We evaluatethe proposed framework on two public spam corpora. The experiment results show that the proposedframework achieves high precision along with low false positive rate.
