Skip to Main Content
The provision of intelligent user assistance has been an ongoing problem in designing computer interfaces. Interactive computing environments must support expert as well as novice users when providing advice for error correction and answers to questions directed to a system. To address these issues, we have investigated the application of fairly well-understood artificial intelligence techniques in novel ways to provide intelligent help. This paper describes the design methodology used to build REASON (Real-time Explanation and SuggestiON), an intelligent user-assistant prototype for a windowed, multitasking environment. REASON's central component is an inference engine that solves problems arising from a user's activity. When the user makes one of several different kinds of errors, the inference engine offers dynamically generated suggestions about what the user might have intended. The user can also query REASON using natural language. In addition to providing suggestions of corrected input or answers to questions, REASON can provide two complementary types of explanations of these responses, derived from the inferences that led to them.
Note: The Institute of Electrical and Electronics Engineers, Incorporated is distributing this Article with permission of the International Business Machines Corporation (IBM) who is the exclusive owner. The recipient of this Article may not assign, sublicense, lease, rent or otherwise transfer, reproduce, prepare derivative works, publicly display or perform, or distribute the Article.