Trustworthy Web Service Selection Based On Social Network

Web services technique allows the development of large-scale applications in open environment, the service provider provides countless web services with similar functions, which makes the selection of web services on the network necessary. However, most service selection mechanism design fixed factor set as the selection criteria, when dealing with some unknown web services, it can not meet the diverse needs of users. In the perspective of social network, both the service provider network and user network are considered, trustworthy service selection framework is proposed. A fuzzy synthetic evaluation method is used to determine the service provider’s synthetic evaluation value, and consumer reputation evaluation is used to determine the trustworthiness of the individual web service. With obtained QoS requirements, QoS data, synthetic evaluation of service provider and the trustworthiness of web services, consumer make decision on the service selection. The main research contents and results of the thesis are summarized as follows:1) The research development of social network in web services were reviewed, the web services concepts and core supporting technique is outlined, trustworthy web service selection method is recalled, some existing problems in web service selection is analyzed, the research progress of web services is summarized, the contents and significance of this thesis is illustrated to support further research.2) The QoS of web services is introduced, the trust evaluation and reputation evaluation of web service selection are described, the importance of social network in web service selection is clarified, the current trustworthy web service selection method and its implementation are summarized, the characteristics and scope of each method is analyzed. These studies lay the theoretical foundation for the establishment of a trustworthy web service selection framework.3) A trustworthy web service selection framework is proposed, trustworthy web service selection is studied in the perspective of consumer network and service provider network, a synthetic evaluation method is used to service provider, information of the consumer network is extracted to determine the trustworthiness of web services, end-users choose the most appropriate QoS of web service. At the same time, the web service selection processes are organized to ensure process standardization. When facing with an unknown service, consumer could obtain the synthetic evaluation of the provider. The autonomic selection of consumers finds solution to the dynamic, personalized and context missing issues in the field of service selection. The combination of consumer network and provider network could fully ensure the trustworthiness and QoS of services.4) Large-scale QoS experimental data sets in the real environment are experimented to verify the accuracy and performance of the service selection framework. This data set contains numerous web service call records, these open web services are randomly selected from the network and it contains150web service consumers and100services. The basic idea is to personalized analyze the QoS value of consumer-service matrix, then we introduce the service selection framework, the obtained QoS values of forecasting trustworthy service are sorted by rules, so that the consumer has the web service with best predictive QoS value. The results of our study show that web service selection results are highly trustworthy, and they can improve the efficiency and performance of service selection in the service selection framework.