基于模糊逻辑的隐私感知Web服务选择研究

发布时间：2018-07-05 11:02

本文选题：Web服务 + 服务选择　；参考：《湖南科技大学》2017年硕士论文

【摘要】：近年来,随着Web服务技术的广泛应用,互联网上出现了大量的Web服务,这些服务通过Web平台为消费者提供各种功能,构成了大规模面向服务计算的基础。由于服务开发者众多,具有相同或相似功能的服务层出不穷,如何在这些相似服务中选择满足用户需求的优质的服务成为了一个关键问题。为此,大量的Web服务发现和选择方法被提出,主要有基于服务质量(Q ua lity of Service,Q oS)的方法、基于信任的方法等。然而,以往的方法很少考虑到用户隐私保护的问题。而在实际的服务选择中,用户大多数情况下需要向服务提供者提供自己的个人数据,如姓名、手机号码等。这些数据可能关系到用户的个人隐私,而目前情况下,用户和服务提供者之间常常缺乏可行的隐私数据保护协议,导致用户在进行服务选择时存在严重的隐私安全隐患。本文工作主要包含以下几个方面:1)针对用户的隐私偏好和服务的隐私策略定义问题,提出了一种柔性的表达方法,允许用户和服务提供者采用自然语言术语描述隐私偏好和隐私策略;2)针对服务选择中的隐私保护问题,不仅考虑了服务信誉的影响,也考虑用户的隐私偏好的影响,通过定义模糊逻辑推理规则把两者结合起来,计算用户在选择不同信誉的服务时所需要付出的隐私代价,然后对候选服务进行排序,将隐私暴露风险最低的服务推荐给用户;3)针对以往服务组合研究的不足,提出了一种隐私感知的服务组合模型,在服务组合中增加了满足用户隐私保护需求的约束条件,把面向服务组合的多目标优化问题转化为最优路径查找问题,在Dijkstra算法的基础上设计算法可以为用户构造隐私代价最低的服务组合。为了证明本文所提出方法的有效性,本文不仅使用了若干例子进行了说明,而且开发了隐私感知的服务选择方法系统原型,同时还利用仿真数据开展了实验,以验证本文方法的性能。
[Abstract]:In recent years, with the wide application of Web services technology, a large number of Web services have emerged on the Internet. These services provide consumers with various functions through the Web platform, which constitute the basis of large-scale service-oriented computing. Due to the large number of service developers and the emergence of services with the same or similar functions, how to select high-quality services to meet the needs of users in these similar services has become a key issue. Therefore, a large number of web service discovery and selection methods have been proposed, mainly based on the quality of service (Q ua lity of Service QoS), trust based methods and so on. However, previous methods rarely consider the protection of user privacy. In the actual service selection, most users need to provide the service provider with their personal data, such as name, mobile phone number and so on. These data may be related to the personal privacy of users, but at present, there is often a lack of feasible privacy data protection protocol between users and service providers, which leads to serious privacy security risks for users in service selection. This paper mainly includes the following aspects: 1) aiming at the problem of user's privacy preference and service privacy policy definition, a flexible expression method is proposed. Users and service providers are allowed to use natural language terms to describe privacy preferences and privacy policies. (2) to address privacy protection issues in service selection, not only the impact of service reputation, but also the impact of users' privacy preferences are considered. By defining fuzzy logic reasoning rules to combine the two, the privacy cost of users when choosing services with different reputation is calculated, and then the candidate services are sorted. This paper proposes a Privacy aware Service composition Model, which can satisfy the user's privacy protection requirements, and puts forward a Privacy aware Service composition Model to overcome the shortcomings of previous researches on service composition. In order to overcome the shortcomings of previous researches, a new service composition model is proposed, which can meet the privacy protection requirements of users. The multi-objective optimization problem of service-oriented composition is transformed into the optimal path finding problem. Based on Dijkstra algorithm, the algorithm is designed to construct the least expensive service composition for users. In order to prove the effectiveness of the proposed method, this paper not only uses several examples to illustrate this method, but also develops a prototype of the privacy-aware service selection method system. At the same time, it also makes use of the simulation data to carry out experiments. To verify the performance of this method.
【学位授予单位】：湖南科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP309;TP393.09

【参考文献】