An ontology enhanced parallel SVM for scalable spam filter training

NEUROCOMPUTING

卷:108

Article

Caruana, Godwin¹  Li, Maozhen^1,2  Liu, Yang¹ 

[1] Brunel Univ, Sch Engn & Design, Uxbridge UB8 3PH, Middx, England

[2] Tongji Univ, Minist Educ, Key Lab Embedded Syst & Serv Comp, Shanghai, Peoples R China

关键词: Spam filtering;    Support vector machine;    Parallel computing;    Classification;    MapReduce;   

DOI  :  10.1016/j.neucom.2012.12.001

来源: Elsevier

【 摘 要 】

Spam, under a variety of shapes and forms, continues to inflict increased damage. Varying approaches including Support Vector Machine (SVM) techniques have been proposed for spam filter training and classification. However, SVM training is a computationally intensive process. This paper presents a MapReduce based parallel SVM algorithm for scalable spam filter training. By distributing, processing and optimizing the subsets of the training data across multiple participating computer nodes, the parallel SVM reduces the training time significantly. Ontology semantics are employed to minimize the impact of accuracy degradation when distributing the training data among a number of SVM classifiers. Experimental results show that ontology based augmentation improves the accuracy level of the parallel SVM beyond the original sequential counterpart. (C) 2012 Elsevier B.V. All rights reserved.

【 授权许可】

Free