【作者】 陈丽；

【导师】 刘洪伟；

【作者基本信息】 广东工业大学 ， 管理科学与工程， 2015， 硕士

【摘要】 目前电子商务信息隐私的研究,要么集中在基于质性研究方法对用户感知隐私保护有效性的研究,要么集中在基于信息技术角度的隐私保护的研究。同时,各种安全保护算法设计仅仅从理论上考虑安全性,但在组织选择决策层面并没有达到预期的效益。此外,传统的组织对隐私信息保护的决策分析研究都是基于决策者为诚实人的假设,然而,在现实情景下,往往极少存在完全诚实的决策者,因此很多研究分析结论并不能有效地与组织存在的隐私信息保护决策问题接轨。因此,研究目标是结合用户个性化服务、隐私保护策略、横向组织间信息共享机制三个研究热点领域,提出电子商务跨平台个性化服务的隐私保护策略和安全信息共享机制,满足跨平台的推荐服务的需要。首先,文章采用信息边界理论、信息敏感度测度论,分析网络信息隐私的边界以及隐私在不同学科不同领域的定义,然后把信息隐私按照0到1的等级划分用户敏感度；其次,采用博弈论、定价模型、效用模型构建两企业之间的静态博弈。观察在自由竞争市场,企业是否实行隐私保护、是否进行个性化服务会对企业受益的影响；再次,采用经济学理论,考虑用户收益、企业利润等,在垄断市场环境下构建两个企业之间的博弈。该博弈模型计算消费者收益函数,平衡企业之间的收益。从中观察企业是否实施隐私保护会吸引何种消费者,同时收益情况如何；最后,采用博弈论、机制设计理论,假设参与者都是理性且不一定诚实,构建效用函数,引入勾结成本的方法,防止参与者之间相互勾结。该机制会降低那些背叛机制的参与者的效益,迫使其改变行为从而获取更高的收益。也就是说当参与者诚实参与机制的时候收益最大,当他们欺骗或者勾结行为的时候,收益反而会下降。最后,在用户敏感度调查中发现用户对于个人身份信息最为敏感,达到0.77；在自由竞争市场中,如果企业实行非对称隐私保护,则可以最大程度的减少价格竞争；在垄断市场中,如果在位者实行隐私保护,而与此同时进入者进入市场,那么两个企业的利润都可以达到均衡。同时消费者的效用也可达到最大化；在两个企业信息共享的机制设计中,引入惩罚机制以后,不诚实的企业则会获得比诚实时候更少的利益。更多还原

【Abstract】 The methods which focus on privacy protection are either based on qualitative resea which is about utility of privacy protection or based on information technology of priv protection. Meanwhile, various algorithms of privacy protection only consider security ft the aspect of theory, but not from the level of organization. In addition, traditic organization analyzes the decision of privacy protection based on the assumption that al the participations are honest. However, in the reality, there barely exist totally honest peo Therefore, many studies which only assume that people are honest cannot effectiv connect with privacy protection.Therefore, the goal of this research is to combine the user personalized service, priv policy with organization mechanism of information, and then to put forward the priv protection of cross-platform personalized service in e-commerce and secure informal sharing mechanism, so that to meet the needs of the cross-platform recommendation serv First we use the theory of information boundary and theory of information sensitivity analyze the boundary of the privacy of network information and the definition of privac; different subject areas. And then we classify the information privacy from 0 to 1 accordin the sensitivity of users.Second we use the game theory, model of pricing and utility to bi up a static game model between two firms. And we figure out whether introducing privacy protection or not and whether offering personalized service or not would affect utility of the firms.Third we use economics theory to build up a dynamic game me between two firms considering benefit of user and utility of enterprise. This model calcu the benefit of users and balance the profit between firms. We figure out whether enterin not and whether offering personalized service or not would affect the utility of firns.Fourth we designed a privacy protection mechanism, prevent collusion betw participants and profit. This mechanism would decrease the benefit of collaborators, wr forces them to be honest. In other words, only the honest people can get a bigger profit.The most sensitive information for customers is ID according to the survey, and sensitiveness is 0.77. In the free competition market, if firms implement asymmetr privacy protection, it can reduce the price competition between two firms. In a monoj market, if the incumbent implements privacy protection while the entrant enters the mat the two companies’ profits can be balanced. At the same time, consumers also can maxin utility. In the design of the mechanism of information sharing between two firms, Disho enterprises will get fewer than honest after introducing punishment.更多还原