电子商务推荐攻击研究
发布时间：2018-07-25 10:23
【摘要】：个性化推荐是实现客户关系管理的重要手段和技术。协同过滤作为最核心、最典型的个性化推荐技术,被广泛应用于电子商务,但其推荐结果对用户偏好信息敏感,使得推荐系统易受到人为攻击,电子商务推荐安全成为个性化推荐能否成功应用的关键。作者先简要介绍了电子商务个性化推荐的基本概念,然后系统阐述了推荐攻击的概念、特征、攻击成本及攻击效率,并详细比较了各种攻击模型,以及各种攻击模型对不同推荐模型的稳定性和健壮性的影响,分析比较了各种攻击检测模型。最后总结评述了电子商务推荐安全的研究现状,并提出了未来研究的挑战。
[Abstract]:Personalized recommendation is an important means and technology for the realization of customer relationship management. As the core and the most typical personalized recommendation technology, collaborative filtering is widely used in e-commerce. However, the recommended results are sensitive to user preference information, making the recommendation system vulnerable to human attack. The recommendation security of e-commerce becomes a personalized recommendation. The key to successful application is to introduce the basic concept of personalized recommendation in e-commerce. Then the concept, characteristics, attack cost and attack efficiency of the recommended attack are systematically expounded, and various attack models are compared in detail, and the effects of various attack models on the stability and robustness of different recommendation models are analyzed and compared. Various attack detection models are presented. Finally, the research status of recommendation security in e-commerce is summarized and the challenges of future research are proposed.
【作者单位】： 中国人民大学信息学院 中国人民大学信息学院 中国人民大学信息学院
【基金】：信息管理与信息经济学教育部重点实验室开放基金资助(F0607-31)
【分类号】：TP393.08
[Abstract]:Personalized recommendation is an important means and technology for the realization of customer relationship management. As the core and the most typical personalized recommendation technology, collaborative filtering is widely used in e-commerce. However, the recommended results are sensitive to user preference information, making the recommendation system vulnerable to human attack. The recommendation security of e-commerce becomes a personalized recommendation. The key to successful application is to introduce the basic concept of personalized recommendation in e-commerce. Then the concept, characteristics, attack cost and attack efficiency of the recommended attack are systematically expounded, and various attack models are compared in detail, and the effects of various attack models on the stability and robustness of different recommendation models are analyzed and compared. Various attack detection models are presented. Finally, the research status of recommendation security in e-commerce is summarized and the challenges of future research are proposed.
【作者单位】： 中国人民大学信息学院 中国人民大学信息学院 中国人民大学信息学院
【基金】：信息管理与信息经济学教育部重点实验室开放基金资助(F0607-31)
【分类号】：TP393.08
【参考文献】
相关期刊论文 前3条
1 赵亮,胡乃静,张守志;个性化推荐算法设计[J];计算机研究与发展;2002年08期
2 余力,刘鲁;电子商务个性化推荐研究[J];计算机集成制造系统;2004年10期
3 余力,刘鲁,罗掌华;我国电子商务推荐策略的比较分析[J];系统工程理论与实践;2004年08期
【共引文献】
相关期刊论文 前10条
1 覃遵跃;在左边是单属性的函数依赖集中寻找关系模式候选码的算法[J];安庆师范学院学报(自然科学版);2003年02期
2 张友志;程玉胜;王一宾;;基于Web日志挖掘的Markov预测模型及算法研究[J];安庆师范学院学报(自然科学版);2010年01期
3 牟乃夏;刘文宝;张灵先;孙翠羽;;空间信息服务的个性化问题[J];测绘科学;2011年03期
4 战坤;曾凡;康运生;戴黎阳;;个性化——医院网站信息服务的趋势[J];重庆医学;2009年21期
5 曹毅;罗新星;;电子商务推荐系统关键技术研究[J];湘南学院学报;2008年05期
6 廖华;;基于客户生命周期与交易偏好的商品推荐方法[J];大家;2010年08期
7 高e，
本文编号：2143524
本文链接：https://www.wllwen.com/guanlilunwen/kehuguanxiguanli/2143524.html
