Sentence Selection using Latent Semantic Analysis for Automatic Question Generation in E-Learning System

Abstract

The Current scenario of the educational system is highly utilizing computer-based technology. For the Teaching-Learning process, both the learners and teachers are highly preferred the online system i.e, E-Learning because of its user-friendly approach such as learning at anytime and anywhere. In the Online educational system, the E-Content plays a major role so the critical importance has to be provided in generating the E-Content. Currently, a large number of study materials are dumped into the internet which has reached the highest limit. The enormous amount of content with high volume leads the learner to skim or frustration in learning. Learners have to spend too much of time to understand their concept from the selected web page. The Tutor also faces the challenges in setting the question paper from this high volume of learning content. We have proposed the computer-assisted system to summarize the learning content of the material using Machine Learning techniques. The Latent Semantic Analysis reduces the size of the content without changing their originality. Finally, the singular value decomposition is used to select the important sentences in order to generate the Multiple Choice Questions (MCQ) to assess the knowledge level of the learner.