Active Learning for Online Spam Filtering

Spam filtering is defined as a task trying to label emails with spam or ham in an online situation. The online feature requires the spam filter has a strong timely generalization and has a high processing speed. Machine learning can be employed to fulfill the two requirements. We propose two kinds ensemble learning methods to combine five simple filters for higher performance. After that we select best four spam filters and add active learning method for choosing training emails. The experiments results show the ensemble learning method has better effects, in which the proposed SVM ensemble method has the highest performance. They also show the filter applying active learning method can reduce requirements of labeled training emails and reach steady-state performance more quickly.

• Type: Conference
• Year: 2008
• Where: AIRS
• Authors: Wuying Liu, Ting Wang
• Keywords: Active Learning Method, AIRS 2008, Ensemble Learning Method, Information Retrieval, Spam Filters