面向个性化云服务的用户行为信任模型研究

发布时间：2018-11-03 12:43

【摘要】：云计算是一种新的“以人为中心”的计算模型，随着用户需求的不断增长，其逐步成为IT领域研究的热点。建立适应于云计算环境的用户行为信任模型有利于提高评估用户和服务提供商之间信任关系的准确性，本文面向个性化云服务，探讨客观公正的用户行为信任模型的建立方法，主要工作如下： 1)为全面评估用户与服务提供商之间的信任关系，并考虑云计算环境下服务交互的动态性和不确定性，提出了一种面向个性化云服务的动态信任模型（PCSDTM模型）。该模型利用灰色系统理论相关知识来反映信任的主观特性，同时建立用户行为的激励/惩罚机制；设计了一种新的基于评价相似度的直接信任与间接信任因子的动态调节方法。实验结果表明，该模型得到的综合信任值更加科学，提高了计算信任值的准确性。 2)分析了云计算环境下用户与服务提供商之间信息交互的整体安全性，针对不同用户类型，建立了面向个性化云服务的隐私保护信任模型（P3Trust模型）。首先，，引入信任评估代理并应用公钥密码技术，构建了一种基于隐私保护的信任评估系统框架，通过保护交互双方如身份标识、兴趣爱好等隐私信息，来抑制合谋欺诈、恶意推荐等恶意行为的发生，使信任评估的结果更加真实；然后，设计了基于用户类型的信任评估方法，调节时间相关性、主观特性、模糊性等相关因子来提高信任评估的准确性及计算效率；提出了一种基于服务质量的信任更新机制，来描述信任的动态性。最后，通过实验验证，该模型能在保护用户隐私信息条件下有效评估用户与服务提供商之间的信任关系，并且可以有效抵御合谋欺诈等恶意行为。
[Abstract]:Cloud computing is a new "human-centered" computing model. With the increasing demand of users, cloud computing has gradually become the focus of research in the field of IT. The establishment of user behavior trust model adapted to cloud computing environment is helpful to improve the accuracy of evaluating the trust relationship between users and service providers. This paper aims at personalized cloud services and discusses the establishment of objective and fair user behavior trust model. The main work is as follows: 1) in order to comprehensively assess the trust relationship between users and service providers, and consider the dynamic and uncertainty of service interaction in cloud computing environment, A dynamic trust model (PCSDTM model) for personalized cloud services is proposed. The model uses the relevant knowledge of grey system theory to reflect the subjective characteristics of trust, and establishes the incentive / punishment mechanism of user behavior, and designs a new dynamic adjustment method of direct and indirect trust factors based on evaluation similarity. The experimental results show that the synthetic trust value obtained by this model is more scientific and the accuracy of calculating trust value is improved. 2) the overall security of information exchange between users and service providers in cloud computing environment is analyzed, and a privacy protection trust model (P3Trust model) for personalized cloud services is established for different user types. Firstly, a trust assessment system framework based on privacy protection is constructed by introducing trust assessment agent and applying public key cryptography technology. By protecting the privacy information of both sides such as identity, interest and hobby, collusion fraud is restrained. The occurrence of malicious behavior such as malicious recommendation makes the results of trust evaluation more true; Then, the trust evaluation method based on user type is designed to improve the accuracy and computational efficiency of trust assessment by adjusting the time correlation, subjective characteristics, fuzziness and other related factors. A QoS-based trust update mechanism is proposed to describe the dynamic nature of trust. Finally, the experimental results show that the model can effectively evaluate the trust relationship between users and service providers under the condition of protecting users' privacy information, and can effectively resist malicious acts such as conspiracy fraud.
【学位授予单位】：安徽师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08

【参考文献】