Combined N-gram and Semantic Approach to Assignment Feedback Analysis and Generation

A recent surge in MOOCs (Massive Open Online Courses) and use of information technology in educationseemed to be a precursor of a disruptive change in education, with a great increase of efficiency in teaching andlearning. Although this increase in efficiency has been hard to achieve, the use of technology has provided amassive increase in collected electronic data consisting of teaching materials, assignments, andinstructor-learner communication. In this project, we explore the use of state-of-the-art natural languageprocessing techniques for analysis of assignment submission content, instructor assessment and feedback, andcomputer-aided instructor support for more efficient and effective feedback. We apply n-gram and semanticbased machine learning techniques for text mining to improve consistency in instructor feedback, and torecommend feedback that will be provided.