Augmenting Online Video Lectures with Topically Relevant Assessment Items

Abstract

In this paper, we present a prototype system for augmenting online video lectures with assessment items generated by analyzing the corresponding text transcripts. A video lecture of longer duration typically covers a number of topics. With linear discourse segmentation approach, we segment a video lecture transcript into topical segments. Inter and intra sentential structures of individual segments are analyzed to generate different types of questions. In this work, the question categories are restricted to factual questions (realized through Multiple Choice Questions-MCQs) and non-factual questions (why, how etc.) that demand higher level cognitive efforts in learner’s part. We have presented evaluation of important modules involved in design of the proposed system. The experimental study has been performed with dataset of 192 video lectures (each having 1-hour duration approximately) covering 5 computer science courses from National Programme on Technology Enhanced Learning (NPTEL) project.