人机界面 agent是一种具有一定智能性 ,并能增强应用系统与用户间交互的计算机程序 .通过机器学习机制 ,这种 agent能够适应用户的习惯 ,使系统更有效的服务于用户 .本文将界面 agent( Interface agent)技术应用于预测支持系统 ( FSS:Forecasting Support System)中 ,并采用适当的学习方法 ,可以很好地解决其中的复杂人机交互问题 The man-machine interface agent is a kind of computer program which has a certain intelligence and can enhance the interaction between the application system and the user. Through the machine learning mechanism, this agent can adapt to the user’s habit and make the system serve the user more effectively. The Interface Agent technology is applied to the Forecasting Support System (FSS) and adopts appropriate learning methods to solve complex human-computer interaction problems