Title
An Intelligent Distributed Environment for Adaptive Learning (IDEAL)
Speaker
James Harner, Chair, Department of Statistics, University of West Virginia
Abstract
IDEAL is a Web-based adaptive learning environment; future enhancements will include an Intelligent Tutoring System (ITS). It consists of the following major components: an XML/MathML content and publishing environment; a statistical modeling subsystem; a mathematical subsystem, and a database repository (e.g., Oracle). IDEAL's core infrastructure consists of Java-based Web applications, servlets, and Java server pages. When an XML document is requested by a learner, it is processed logically using the student's information stored in the backend database before being returned. The statistical subsystem, JavaStat, is a Java application which runs on the server and uses R as a backend statistical engine. R is used for advanced modeling and will be the engine for the learning theory models discussed below. The statistical subsystem will also be used for teaching statistics. JavaStat drives a large number of instructional applets and allows the instructor to conduct collaborative sessions. The mathematics subsystem, JavaMath, is similar in design to JavaStat and will be used for remedial algebra and geometry sessions and it will drive the statistical distribution applets. By design, IDEAL's architecture is content independent, but it is structured towards problem solving subject areas. The principal parts of IDEAL that directly support learning are the tutorial examples and exercises. Tutorial Examples embed a series of questions within the document's hierarchical structure and the content is revealed to the student sequentially based on his/her answers. Cognitive assessment models are being developed as part of a mastery learning strategy for these tutorial examples. Exercises (and exams) are used for evaluating the student's learning outcomes, which will be assessed by item response theory (IRT) models.
