以數(shù)據(jù)分析平臺(tái)的建設(shè)為背景，深入分析了數(shù)據(jù)分析平臺(tái)的特點(diǎn)和需求，在此基礎(chǔ)上，結(jié)合Agent技術(shù)，提出了具有智能化、個(gè)性化、平臺(tái)化等特點(diǎn)的數(shù)據(jù)分析平臺(tái)多Agent 系統(tǒng)的模型，研究了模型中Agent 間的通信與交互方式，采用點(diǎn)對(duì)點(diǎn)的消息傳遞的方式進(jìn)行Agent 間的通信，采用FIPA-ACL 通信語(yǔ)用標(biāo)準(zhǔn)，以數(shù)據(jù)源選擇為例對(duì)Agent 間的通信過(guò)程進(jìn)行了設(shè)計(jì)，為數(shù)據(jù)分析工具的研究與開發(fā)提供了一定
的借鑒。
關(guān)鍵詞：數(shù)據(jù)分析平臺(tái)；多智能體；Agent 通信；信息系統(tǒng)
Abstract: Based on the construction of the data analysis platform the characteristics and requirements of the data analysis platform are analyzed. The model of multi-agent system for data analysis platform is proposed that makes the work of data analysis intelligent, individualized and facilitate by using the agent technology. The communication mechanism is studied among the different agents by using point-to-point message passing and
FIPA-ACL. Taking the selection of data sources as an example, the communicating process among the agents is designed. The research work is a reference to the research and development of the data analysis platform.
Key words: data analysis platform; multi-agent; agent communication; information systems