基于贝叶斯网络的多Agent服务推荐机制研究-中国科学院.PDFVIP

37 4 Vol. 37 No.4 2010 4 Computer Science Apr 2010 Agent ( 中国科学院计算技术研究所 北京100190) 普适计算环境中的服务推荐需要满足系统异构性和移动性的要求提出了一种基 贝叶斯网络的多Agent 服务推荐机制并进行实现, 将贝叶斯网络和聚类方法应用 服务推荐中, 并设计了推荐模型自学习机制, 充分考虑了 上下文对服务推荐的影响及改进实现系统由完成历史上下文汇集知识训练决策推荐和自学习功能的多 个Agent 构成, 通过 Agent 之间的通信内容设计,在 Agent 之间建立流程控制和数据共享通道 普适计算, 服务推荐, 多Agent, 贝叶斯网络 TP39 A Research on MultiAgent Service Recommendation Mechanism Based on Bayesian Network DU Jing YE Jian SHI Hongzhou HE he H U henmin ( Institute of Computing T echnology, Chinese Academy of Sciences, Beijing 100190, China) Abstract In ubiquitous computing environment, services recommendation needs to satisfy the system s mobility and he terogeneity. A multiAgent service recommendation mechanism was put forward, bayesian network and clustering meth od and update mechanism w ere used.It is fully taken into account that contexts affect t he result of recommendation.The system consists of Agents w hich accomplish the funct ions such as hist ory cont ext collecting, knowledge training, policy making recommendat ion and selflearning. Furthermore, by designing the information of communicat ion, flow control and data sharedpath among Agents were constructed. Keywords Ubiquitous computing, Service recommendation, Multi Agent, Bayesian netw ork 1 Annie Chen , , [ 3] , 3 ∀ ∀ ∀ 4 ∀ ∀ ∀ , ! ∀ ∀ ∀ Annie Chen