| With the rapid development of cloud computing, social networking and other emerging technologies, human society has entered a new information platform era. Cloud platform provides a variety of features and services to meet the needs of people working and living. With the popularity of mobile intelligent terminal, the user’s personal data is stored in the cloud, and data processing is also done by the cloud service. Popularity of intelligent terminals also brought prosperity of social networks, people can share thoughts and activities at any time in order to maintain and expand their communication groups. Using cloud platform information management and releasing by the social network is a modern way of life. However, the open nature of cloud platforms and social networks lead to data security, privacy leaks and other problems, how to make it safe when enjoying the convenience of modern technology without having to worry about damage to their own interests is an urgent problem. Trust is the key to solving this problem. In this paper, we studied and summarized the existing trust evaluation program in the Cloud environment and Social networks, then we propose the appropriate models for different background. The author’s major contributions are outlined as follows:(1) In social networks, the existing trust evaluation model only focus on the experience or behavior pattern. These models lead to one-sided result. To solve this problem, we proposed an evaluation model which adopted of user’s attitudes, interactive experience, and behavior patterns comprehensively. Experimental results show that our model can be adapted to the dynamics and complexity of the social network. Compared to existing methods, the model has higher accuracy.(2) A novel indirect trust evaluation model for multiple context of social networks is proposed to deal with the problem that the context sensitivity of trust lead to the trust evaluation challenge between indirect users. This model takes advantage of the relevant concepts, through the comprehensive analysis of network structures and user’s trust relationship in each context. The model establishes relevance network on top of trust networks, then uses the relevance of context compute user’s trust across the context. It avoids the effective of the trust attenuation and trust path between indirect users is difficult to find in multiple context and sparse networks, targeted build relevance networks of user groups, ensuring evaluation accuracy and reasonableness. Experimental results of real social networks shows that our model not only can compute the indirect trust value in one context, but also more suitable for the multiple context. Compare with the existing model to evaluation, its accuracy has greatly improved.(3) We studied service composition problem from the perspective of network, and used a method which guarantee the connectivity of entire service network to ensure the functions. We use the totally experience based trust evaluation model to compute the trust value of nodes in the service network. Then, we use the method of analyzing the robustness of networks in random graph, which guarantees conditions for large-scale service environment to ensure the trusted service composition.(4) In the cloud environment, service consumers lack experiences of interaction and malicious adviser will lead to unfair selection. In this paper,we propose a two-layer selecting model to meet the demand of consumer. In the first layer consumer use social records update owns to get more accurate experience. Afterwards, the second layer, consumer select the most suit adviser. The result shows that our approach can filter the malicious participants, and choose the services which fit consumer’s preference.
|
PDF Full Text Request