The University of Edinburgh -
Division of Informatics
Forrest Hill & 80 South Bridge

MSc Thesis #9573

Title:Course Selector: a System to Support Course Selection in the Department of Artificial Intelligence
Date: 1995
Abstract:Artificial Intelligence Course Selector (AIS) is a forward chaining rule based system in CLIPS 6.0. It extends an earlier system, Course Selector (CS), that helps students choose course combinations in the Business Studies Department of the University of Edinburgh (Kingston 1992). Normally this task is completed by reading the University Calendar (an 800 page volume) to create a course schedule that conforms to University regulations and time-table restrictions. However students often rely upon their Director of Studies for this expertise and consume large amounts of his or her time. This project takes the CS system and transfers it for use by students in the Department of Artificial Intelligence (AICS). In addition new functionality has been added to the system to augment the task of selecting courses. This takes the form of data driven algorithms to reason about the future consequences of course selections. Also the system has been modified to allow more flexible use and access to the knowledge. Evaluation was conducted in four stages. Firstly, the rule base was checked for consistency. Secondly the system was tested using cornerstone examples of the domain. Thirdly the system was evaluated by the Director of Studies for the Department of Artificial Intelligence. Finally, the system was evaluated by the end users. Overall the results were positive and the system functioned correctly during these trials.

[Search These Pages] [DAI Home Page] [Comment]