Adaptive Intelligent Tutoring Systems

This thesis descnbes the research and development into the intelligent tutoring system called JASMINE (JAVA Adaptive Student Modelling In a NLP Environment). JASMINE is designed to address the ideas of adaptive and responsive, presentation and remediation in tutonng. Remediation is indicated when the confidence level of a given fragment of knowledge falls below a set mastery level. As a result, the student is presented with tirnely assistance almg with presentation adapted to the students preferred style of learning. Ln order to achieve this, the model of Neuro-Linguistic Programming (NLP) was considered, with the main input channe1 of the model (visual, auditive or kinetic), identified for each student. The approach withh JASMINE is to forecast the students most highly valued input channel through the use of a number of NLP related questions during the initial session. Then, to use monitoring techniques and to adjust the system's presentation to the student. To aid in modelling and knowledge storage, the User Modelling shell system BGP-MS was investigated and found lacking, paNcularly in the ability to store student knowledge Ievels and provide a cross platform environment. Hence, JAVA'S role within J A S m grew from the initial intention of interface integration to the irnplementation of the complete student model. This allowed the JASMINE application to become a prapatic solution to the ITS question.