Research On Trust Evaluation Model And Application Based On Cloud Theory

Academician Li Deyi put forward the cloud theory whose core is cloud model. The cloud model is a strong new method to achieve the transition between the qualitative concept and quantitative values, which overlap probability theory and fuzzy set theory through a specific structure of the crossover operator. It can provide much valuable information than other mathematic method when used to expressing information with cloud model can provide more valuable information than with the other mathematical methods, being identically express the subjective trust in the existence of randomness, fuzziness, and the relationship between the two. Trust evaluation model for the study of trust is not fully take into account the uncertainty in general, so to carry out the research on trust evaluation model based on cloud model is very significant.At first, the main tool of cloud model is introduced into the model for measuring the evaluation entity of its operation environment and transaction experience. And a trust cloud evaluation model is designed based on two-dimension measuring. The model classifies the trust into operation environment trust and transaction trust, and the cloud model and feedback evaluation method are adopted respectively to evaluate them from two aspects. In the calculation of operating environment'trust value, the calculation method of trust value of entities'operating environment based on Trust Cloud is given, and in the trust decision method, the entity's average trust situation is reflected by the parameter Ex, the uncertainty of trust is reflected by the other two parameters. The expression can retain the inherent uncertainty of trust most, so it has greater advantage than a single data in expressing trust information. In the calculation of transaction trust value, the calculation method of trust value of recommendation transaction based on Trust Cloud is given in the text, achieving the purpose of effectively distinguishing the recommendations entities of the malicious unstable behavior. And the tragedy for choosing transaction entities based on Trust Cloud is given, not only achieve the purpose for giving priority to some entities of good reputation and stable behavior to choose,but also effectively solve the problem about how to choose the entity who having the same trust degree. At the same time, the constructed model is applied for the power supply enterprise's trust assessment to large customers, making the assessment more objective and accurate. It provides a decision support method for power supply companies choosing reputable large power customers.In addition, for the trust updating issue in the trust network, a Bayesian feedback trust cloud model is designed. The model is composed of the cloud droplet generation module and cloud droplets test module and the Bayesian feedback module. Among them, the cloud droplet generation module is used to generate the set of cloud droplets. Main function of the cloud droplet test module is used to conduct statistical tests of cloud droplets and to remove the deviated cloud droplets from the set of cloud droplets, and then collect some new samples. Bayesian Feedback module is based on Bayesian theory, and feedback correction is made into initial trust cloud parameters by using the new samples parameters. Finally, the two modules were carried out after the principles of analysis and system implementation. The new cloud parameters can better reflect the current people's understanding of the entities, and will not ignore the initial expert's opinions. Information fusion is realized in the model, and the updating issue of trust information has resolved.