TEACHMET: An Expert System for Teaching Weather Forecasting

Abstract

Students or weather forecasting need to learn to identify efficiently the information relevant to the elements they predict. One way students learn these skills is by engaging in discussions of interesting weather situations as they occur. In these discussions, the mentor explains his reasoning in developing the forecast, and his rationale for using particular data. This process is very time consuming and relies on the student's willingness to ask questions. In this project, we developed a computer-assisted instructional system that interactively guides the student of weather forecasting in choosing and interpreting the wide variety of available data. The system is capable of giving detailed responses to the student's questions and is available whenever the student wishes to use it. The system incorporates the expertise of several faculty members and graduate students, as well as two National Weather Service forecasters. A description is given of the creation and evaluation of the system as well as difficulties that were encountered.