科研之友
微信
新浪微博
Facebook
分享链接
摘要
With the development of network infrastructures and the Internet, using web services to access and process information in distributed form apart from the different hardware and software platforms using standard protocols, Internet data exchange is expanded. By using web services, distributed applications simply exchange data, or do things by working together. In web-based applications, trust is one of the criteria for selecting the appropriate web service and also web service criteria are often qualitative. Because of the importance of this issue, we used fuzzy system in this article to assess the trust of single web services. To assess the trust of web services, eight parameters includes QoS, user preferences, subjective calculation of trust, objective calculation of trust, bootstrapping, dynamics of trust and independent trust have been considered in proposed fuzzy system. To obtain a fuzzy system with high estimation accuracy, we considered 24 different fuzzy systems with different fuzzy sets and deffuzication methods and then we created fuzzy sets and fuzzy rules based on training data. To increase the speed and to reduce the number of fuzzy rules, we have considered the fuzzy systems as three-level. Finally, after running fuzzy systems with test data, between these systems, we received the following results: best fuzzy system achieves root mean square error 2.27 percent, the accuracy value 0.9816 and the FScore value 0.9838 and recalls value 0.9860.
- 出版日期2017
