The Research On Trust Evaluation Model In Web Services Environments

As is well known, Web services as a new kind of business application pattern, it is a developmental result of distributed computing middleware model, and the safety issue has become a critical factor of restricting its development. Trust evaluation model research provides a way to solve the problem of Web services safety. Trust evaluation model means it establishes and manages trust relationships framework that is the premise and guideline of the platform framework design which is the important part of trust evaluation mechanism. Thus, constructing a rational and effective trust evaluation model is the key to reduce the interaction risk of Web services under the Web services environments which are full of openness, dynamic and deceptive. So, we establish two kinds of trust evaluation model which are used for Web services with different mathematical tools. Both of them can evaluate effectively the trust relationship between service entities, and make a rational decision. So they effectively improve the success rate of Web services interaction and ensure the interactive security of Web services. In the research of Web services trust evaluation model, we have done some researches as follows:①A comprehensive survey of trust models in Web services environments was made. The newly-proposed models were analyzed, classified and discussed.②The existing trust evaluation model has been studied, analyzed the relationship of trust in their assessment, description and calculation of the strengths and weaknesses, as well as the weaknesses which applied it to Web services environment, it proposed the basic requirements of trust model design and summarizes the trust evaluation basic principles of entity behaviors.③A Web services trust evaluation model which based on the fuzzy Petri net was proposed. The model combined a graphical tool of fuzzy Petri net with evaluate trust relationship. It made the trust inference analysis more intuitive which liked a traditional flow chart and more in keeping with thinking and cognitive style of human. In this process, it used completely matrix operations; it fully reflected the parallel reasoning abilities of fuzzy Petri net. It not only ensured the security of entity trust information, but also reduced the complexity of entity trust calculation. Through this simulated experiment, we validated the capability of the model and its advantages which compared with other models.④A Web services trust evaluation model which based on the fuzzy theory was proposed. The model used fuzzy theory to re-define the crucial links of trust evaluation procedure that made it more specific; it improved the practicality and validity of the model. Using fuzzy combination weight approach could reasonably distribute the weight of trust evaluation factors set and solved the accuracy of service trust feedback evaluation. The concept of entity evaluation trust was introduced, the actual recommended evaluation ability of entity of effectively response. The traditional UDDI and WSDL model have expanded, increasing the description and evaluation mechanisms of trust; it achieved publication, selection and sorting mechanism of Web services resource which based on trust relationship. We verified the effectivity and practicability of the trust model after doing the simulated experiment.In the last part of dissertation, several issues for further in-depth research and exploration are given.