信任、声誉等基于社群的自适应性测度已成为网格环境下支撑服务质量需求的砖石.借鉴社会学、心理学和诸如P2P计算、普适计算等计算机科学中的相关研究,网格环境下的信任声誉研究主要应用在资源管理和调度中.对网格环境信任声誉评估的典型模型进行评述与比较,包括基本模型、离散模型、Bayesian模型、信念模型和模糊逻辑模型.在设计评估模型时应该考虑计算收敛能力,动态性和健壮性等多方面因素.考虑评估模型对负载的影响,设计针对网格环境特性的信任声誉评估模型是一个发展方向.
Community - based adaptive metrics such as trust and reputation serve as building blocks to support quality- of- service requirements in Grid environment. Borrowed from related research in sociology, psychology and computer science such as P2P computing and Ubiquitous computing, trust and reputation research mainly used in Resource Manage and Schedule in Grid computing. Several typical trust computing models including basic model, discrete model, Bayesian model, belief model etc, are described in detail. Advantage and disadvantage of these models are analyzed and compared. Convergence, dynamics and robust adaptation should be taken into account. Designing a focused trust and reputation evaluation model in Grid computing, which considering effect on load, is a research trend.