Input Data Management in Real-Time AI Systems

A real-time AI system in the real world needs to monitor an immense volume of data. To do this, the system must filter out much of the incoming data. However, it must remain re­ sponsive to important or unexpected events in the data. This paper describes some simple ap­ proaches to data management, shows how they can fail to be both adequately selective and re­ sponsive, and presents an approach that im­ proves on the simple approaches by making use of information about the system's resources and ongoing tasks. The new approach has been ap­ plied in a system for monitoring patients in a surgical intensive-care unit. 1 I n t r o d u c t i o n When an AI system meets the real world, it is confronted by an overwhelming amount of data. To maintain real­ time performance, the system must reduce the flow of data to a manageable amount. Some simple approaches have been used to filter the incoming data, but these have severe shortcomings. To operate in real time while remaining res...

• Type: Conference
• Year: 1989
• Where: IJCAI
• Authors: Richard Washington, Barbara Hayes-Roth
• Keywords: Data Stream, IJCAI 1989, IJCAI 2007, Incoming Data, Real World