Using Multi-Phase Cost-Sensitive Learning to Filtering Spam

摘要：

This paper proposes a novel ensemble learning framework, namely, Multiple-Phase Cost- Sensitive Ensemble Learning (MPCSL) which simulates the means and process of human being learning. It consists of two types learning, i.e., direct learning which learns multiple weak learners from a training dataset via some homogeneous or heterogenous algorithms, and indirect learning that constructs a committee from the knowledge of the combined filters or other committees. This paper studies empirically the performance of MPCSL on spam filtering tasks. In the occasions of combining homogeneous and heterogenous, how the performance of MPCSL changes is surveyed. The results shows that MPCSL is a compellent ensemble learning method for cost-sensitive tasks such as spam filtering.

关键词： machine learning text processing spam filtering cost sensitive learning

作者: Wenbin Li Yiying Cheng TaiFeng Liu Xindong Zhang Ning Zhong

作者单位: School of Information Engineering Shijiazhuang University of Economics, Shijiazhuang, Hebei, China 0 School of Information Engineering Shijiazhuang University of Economics, Shijiazhuang, Hebei, China 0 International WIC Institution Beijing University of Technology, Beijing, China 100022

会议类型: 国际会议

会议名称: 2009 Ninth International Conference on Hybrid Intelligent Systems(第九届混合智能系统国际会议 HIS 2009)

会议地点: 沈阳

会议语种:英文

页码: 1-6

在线出版日期: 2009-08-12（万方平台首次上网日期，不代表论文的发表时间）