Intelligent tutoring system based on semantic multimedia browsing and retrieval

Abstract

An Intelligent tutoring system simulates the functions of a human tutor, such as reproducing the lecture, explaining the difficulties, and interacting with the students. This paper proposes a web-enabled intelligent tutor for our human-centered multimedia e-learning system, which includes an online virtual classroom and offline self-review web. To reproduce the lecture, we provide a multi-level browsing with slides, web, handwriting, audio and video based on automatic multimedia segmentation and reconstruction. For the review of the difficult parts, visual content based retrieval and course content based retrieval are proposed respectively. We integrate efficient interaction in every part of our e-learning system. This system has proved helpful in enhancing the engagement and extent of understanding under the real teaching environment in our university.