An Architecture of Domain Independent and Extensible Intelligent Tutoring System based on Concept Dependencies and Subject Paths

Abstract

Intelligent Tutoring Systems (ITS) seek to provide personalized tutoring to learners, but are often domain specific, and lack extensibility. When featuring extensibility and domain independence, it is a challenge to provide appropriate level of personalization for every learner. In this paper, an architecture of a system that features domain-independence and extensibility with personalization and automatic course improvements without requiring persistent subject expert intervention has been proposed. The proposed architecture utilizes the notion of concept dependencies and the ability to sequence inter-dependent concepts intelligently into subject paths that enable automated tutoring as well as effective course customization per learner. It features a separate interface for subject experts through which they do not require ITS building knowledge to fulfil their appropriately assigned tasks assisted intelligently by the system, and an API based interface layer that supports today’s mobile requirements for better engagement.