基于代价敏感支持向量机的推荐系统托攻击检测方法

J4 ›› 2014, Vol. 36 ›› Issue (04): 697-701.

基于代价敏感支持向量机的推荐系统托攻击检测方法

吕成戍

（东北财经大学管理科学与工程学院，辽宁大连116025）

收稿日期:2012-11-29 修回日期:2013-01-10 出版日期:2014-04-25 发布日期:2014-04-25
基金资助:
辽宁省社会科学规划基金资助项目（L10BJL035）；中央高校专项科研基金资助项目（DUT10RW302）

Shilling attack detection method of recommender
systems based on costsensitive SVM

LvChengshu

（School of Management Science and Engineering,Dongbei University of Finance and Economics,Dalian 116025,China）

Received:2012-11-29 Revised:2013-01-10 Online:2014-04-25 Published:2014-04-25

摘要/Abstract

摘要：

基于标准支持向量机的托攻击检测方法不能体现由于用户误分代价不同对分类效果带来的影响，提出了一种基于代价敏感支持向量机的托攻击检测新方法，该方法在代价敏感性学习机制下引入支持向量机作为分类工具，对支持向量机输出进行后验概率建模，建立了基于类别隶属度的动态代价函数，更准确地反映不同样本的分类代价，在此基础上设计了代价敏感支持向量机分类器。将该分类器应用在推荐系统托攻击检测中，并与标准的支持向量机方法、代价敏感支持向量机方法进行比较，实验结果表明，本方法可以更精确地控制代价敏感性，进一步提高对攻击用户的检测精度，降低总体的误分类代价。

关键词: 支持向量机, 托攻击检测, 代价敏感, 类别隶属度

Abstract:

The shilling attack detection method based on traditional Support Vector Machine (SVM) can not reflect the influence of the great difference of misclassification cost on classification effect. The paper proposes a novel shilling attack detection method based on costsensitive SVM. The method uses SVM as the classification tool under the costsensitive learning mechanism, and models the probabilistic outputs of SVM, presents dynamic misclassification cost function based on class membership value, thus accurately reflecting the misclassification cost of different samples. Hence, we design the classifier of cost-sensitive SVM and apply it in shilling attacks detection. Experimental results show that, compared with traditional SVM and different classes of costsensitive SVM (CSSVM), our proposed method can more precisely control the cost of misclassification, improve the precision of the minority class samples, and decrease the total cost of misclassification.

Key words: support vector machine (SVM);shilling attacks detection;costsensitive;class membership

吕成戍. 基于代价敏感支持向量机的推荐系统托攻击检测方法[J]. J4, 2014, 36(04): 697-701.

LvChengshu. Shilling attack detection method of recommender
systems based on costsensitive SVM [J]. J4, 2014, 36(04): 697-701.

参考文献

［1］Zhang Jing, He Famei, Qiu Yun. Inspection method of the attack on personalized recommendation system description file［J］. Journal of University of Electronic Science and Technology,2011, 40(2):250254. (in Chinese)
［2］Vapnik V N.Statistical learning theory［M］.New York:Wiley,1998.
［3］Vapnik V N .The mature of statistical learning theory［M］.Germany:Springer,2000.
［4］Williams C A，Mobasher B，Burke R D.Defending recommender systems:Detection of profile injection attacks［J］.Service Oriented Computing and Applications,2007,1(3):157170.
［5］Li Cong, Luo Zhigang, Shi Jinlong. An unsupervised algorithm for detecting shilling attacks on recommender systems［J］.Acta Automatica Sinica,2011,37(2):160167.(in Chinese)
［6］Breiman L, Friedman J H, OlshenR A, et al. Classification and regression trees［M］. California:Wadsworth International Group,1984.
［7］Veropoulos K,Campbell C,Cristianini N.Controlling the sensitivity of support vector machines［C］∥Proc of the International Joint Conference on AI(IJCAI’1999), 1999:5560.
［8］Burges C. A tutorial on support vector machines from pattern recognition［J］.Data Mining and Knowledge Discovery,1998, 13(2):121167.
［9］Paola C, Elena C, Giorgio V. Support vector machines for candidate modules classification ［J］.Neurocomputing,2005,68(4):281288.
［10］Japkowicz N.Conceptlearning in the presence of betweenclass and withinclass imbalances［C］∥Proc of CSCSI’01, 2001:6777.
［11］Platt J C. Probabilistic output for support vector machine and comparisons to regularized likelihood methods［M］. Boston:MIT Press,1999.
［12］http://www.grouplens.org/node/73.
［13］Burke R，Mobasher B，Williams C. Classification features for attack detection in collaborative recommender systems［C］∥Proc of ICKDDM’06,2006:156167.
附中文参考文献：
［1］张靖，何发镁，邱云. 个性化推荐系统描述文件攻击检测方法［J］. 电子科技大学学报,2011, 40(2):250254.
［5］李聪，骆志刚，石金龙.一种探测推荐系统托攻击的无监督方法［J］.自动化学报, 2011, 37(2):160167.

[1]	刘屹成, 刘晓燕, 严馨. 并行平衡级联支持向量机[J]. 计算机工程与科学, 2023, 45(07): 1170-1177.
[2]	唐宇, 代琪, 杨梦园, 陈丽芳, . 改进麻雀搜索算法优化SVM的异常点检测[J]. 计算机工程与科学, 2023, 45(02): 346-354.
[3]	仲臣, 余学祥, 邰晓曼, 韩雨辰, 肖星星, 刘清华, . 萤火虫算法优化支持向量机室内定位研究[J]. 计算机工程与科学, 2022, 44(11): 1968-1975.
[4]	马汉达, 朱敏. 改进SVM不平衡数据分类的IGWOSMOTE方法[J]. 计算机工程与科学, 2022, 44(06): 1133-1140.
[5]	贾俊杰, 段超强. 基于评分离散度的托攻击检测算法[J]. 计算机工程与科学, 2022, 44(03): 554-562.
[6]	陈丽芳, 代琪, 赵佳亮. 不平衡数据多粒度集成分类算法研究[J]. 计算机工程与科学, 2021, 43(05): 917-925.
[7]	曹婷婷, 张忠林. 代价敏感的KPCA-Stacking不均衡数据分类算法[J]. 计算机工程与科学, 2021, 43(03): 525-533.
[8]	韦修喜1，黄华娟1，周永权1,2. 基于AP聚类的约简孪生支持向量机快速分类算法[J]. 计算机工程与科学, 2019, 41(10): 1899-1904.
[9]	王云艳1,2，罗冷坤1，王重阳1. 基于流形学习的光学遥感图像分类[J]. 计算机工程与科学, 2019, 41(07): 1212-1219.
[10]	冯璋，裴东,王维. 基于改进灰狼算法优化支持向量机的人脸识别[J]. 计算机工程与科学, 2019, 41(06): 1057-1063.
[11]	崔建双，车梦然. 基于多分类支持向量机的优化算法智能推荐系统与实证分析[J]. 计算机工程与科学, 2019, 41(01): 153-160.
[12]	田勋，汪西莉. 基于聚类标签均值的半监督支持向量机[J]. 计算机工程与科学, 2018, 40(12): 2265-2272.
[13]	颜学龙，龚流青，汪斌斌. 基于DCQGA-SMKL-SVM的模拟电路故障诊断方法[J]. 计算机工程与科学, 2018, 40(11): 1944-1950.
[14]	张和杰,马维华. 基于贴近度等级的链路质量评估方法[J]. 计算机工程与科学, 2018, 40(11): 1991-1999.
[15]	王志华,罗齐,刘绍廷. 基于混沌灰狼优化算法的SVM分类器研究[J]. 计算机工程与科学, 2018, 40(11): 2040-2046.