Developing Student Model for Intelligent Tutoring System

Purushothman Ravichandran, (2018) Developing Student Model for Intelligent Tutoring System. PhD thesis, Open University Malaysia.

Preview	Text (Chapter 1 and Reference only) library-document-1213.pdf - Submitted Version Download (927Kb) | Preview
	Text (Full) library-document-1213.pdf - Submitted Version Restricted to Repository staff only Download (4Mb) | Request a copy

Abstract

The effectiveness of an e-learning environment mainly encompasses on how efficiently the tutor presents the learning content to the candidate based on their learning capability. It is therefore inevitable for the teaching community to understand the learning style of their students and to cater for the needs of their students. One such system that can cater to the needs of the students is the Intelligent Tutoring System (ITS). To overcome the challenges faced by the teachers and to cater to the needs of their students, e-learning experts in recent times have focused in Intelligent Tutoring System (ITS). There is sufficient literature that suggested that meaningful, constructive and adaptive feedback is the essential feature of ITSs, and it is such feedback that helps students achieve strong learning gains. At the same time, in an ITS, it is the student model that plays a main role in planning the training path, supplying feedback information to the pedagogical module of the system. Added to it, the student model is the preliminary component, which stores the information to the specific individual learner. In this study, Multiple-choice questions (MCQs) was administered to capture the student ability with respect to three levels of difficulty, namely, low, medium and high in Physics domain to train the neural network. Further, neural network and psychometric analysis were used for understanding the student characteristic and determining the student’s classification with respect to their ability. Thus, this study focused on developing a student model by using the Multiple-Choice Questions (MCQ) for integrating it with an ITS by applying the neural network and psychometric analysis. The findings of this research showed that even though the linear regression between real test scores and that of the Final exam scores were marginally weak (37%), still the success of the student classification to the extent of 80 percent (79.8%) makes this student model a good fit for clustering students in groups according to their common characteristics. This finding is in line with that of the findings discussed in the literature review of this study. Further, the outcome of this research is most likely to generate a new dimension for cluster based student modelling approaches for an online learning environment that uses aptitude tests (MCQ’s) for learners using ITS. The use of psychometric analysis and neural network for student classification makes this study unique towards the development of a new student model for ITS in supporting online learning. Therefore, the student model developed in this study seems to be a good model fit for all those who wish to infuse aptitude test based student modelling approach in an ITS system for an online learning environment. (Abstract by Author)

Actions (login required)

View Item