基于本体和会话策略的多主体交互协议描述方法  

Specification of multi-agent interaction protocols based on ontology and conversation policy

作　　者：姚永雷[1] 马利[1] 

机构地区：[1]南京信息工程大学计算机与软件学院,江苏南京210044

出　　处：《电子技术应用》2009年第2期101-103,107,共4页Application of Electronic Technique

摘　　要：提出一种基于会话策略的多主体交互协议描述方法。交互协议中的消息用言语动作来表示,这些言语动作被描述为WS-Agreement的schema;会话策略则描述了消息传递的流程以及交互过程中的上下文信息,如参与者属性、时间阈值等等,所有这些会话策略组成了一个多主体交互协议;采用本体描述语言OWL作为会话策略的表示语言。这种方法使得主体在一个开放、动态的环境中可以灵活地选择交互协议。This paper presents a conversation policy based method to specify multi-agent interaction protocols. Specifically, a multi-agent interaction protocol includes speech-acts as the individual messages, and such speech-acts are modeled as WS-Agree- ment schemas. Then conversation policies define the order in which speech acts can be performed and specify the context information of the interaction process. All of these conversation policies collectively describe an interaction protocol. These conversation policies are expressed by the ontology language, OWL. This enables the agents to choose protocols flexibly in a dynamic and open environment.

关 键 词：多主体系统 交互协议 本体 会话策略 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]