Intelligent tutoring system: an expert-system trainer for herring roe grading

Describes an expert system developed to train humans to grade herring roe. The design features of the system are given. The system provides a user with some general information about the herring roe industry, how roe is processed and graded in an industrial plant, and an intelligent tutoring module. The knowledge was extracted from a human expert who is highly experienced in grading herring roe. The system runs under a convenient tool used to develop the expert system. An important criterion in the design of the grading trainer was user-friendliness, since most users are unlikely to be highly educated. To assess the progress of each user, an evaluation module is provided. A module to evaluate the reaction of a user to the system itself automatically validates the goal of the system. Based on the outcome of these modules, final comments are given to the user about his/her progress with the knowledge domain. In designing the training tutorial, different methods of associating information were used: exclusive sets, multi-choice sets, and fuzzy sets. Various types of questions are considered in training the user as well as simulating the realistic condition of grading in an industrial plant where each decision must be made in a fraction of a second. An example is given to illustrate the effectiveness of the system, by evaluating the grading capability of an inexperienced human grader before and after using the intelligent tutoring system.