论文部分内容阅读
反馈信息不能反映评价人的个人偏好,且其评价的重要程度随时间增加逐渐消减。为此,在充分考虑评价人的个人偏好基础上提出了一个基于信誉的信任管理框架。该框架主要包括基于决策树的信任计算模型和时间衰减计算模型。最终使得云服务的信誉值不再仅仅局限于反映云服务质量的一个客观数值,而是更能体现出不同云服务评价人的主观性。在该模型的基础上,进行了实验对比分析。实验分析结果表明,与现有的信誉评价模型相比,采用决策树和时间衰减模型信誉评价更为精确。
The feedback information does not reflect the individual’s preference of the evaluator, and the importance of his evaluation diminishes over time. Therefore, based on the personal preference of the appraiser, a credit management framework based on credibility is proposed. The framework mainly includes decision tree-based trust computing model and time decay computing model. Finally, the credit value of the cloud service is no longer limited to reflect an objective value of cloud service quality, but better reflect the subjectivity of different cloud service evaluators. On the basis of the model, experiments were carried out comparative analysis. Experimental results show that, compared with the existing credit evaluation model, credit rating using the decision tree and the time decay model is more accurate.