An Adaptive Interface for Computer-Assisted Rubrics in an E-Marking Tool Using Nearest Neighbor

Abstract

With an always growing number of student enrolment in higher education, providing quality feedback in both digital and paper based assessment becomes a heavy burden for teachers and tutors. The use of assessment rubrics can help overcome this burden, defining several criteria including formative feedback within it. However, current computer-assisted rubric systems present some drawbacks that hinder their adoption. This paper presents the design of an adaptive interface for an e-Marking tool that uses rubrics, that is able to suggest assessors on the next criterion to be evaluated, based on nearest neighbor approach. An experimental setup showed encouraging results that provide evidence that the use of advanced learning technologies for assessment can help improve efficiency even in ill-defined domains.