An approach to the automatic grading of imprecise diagrams

Thomas, Pete; Smith, Neil and Waugh, Kevin (2006). An approach to the automatic grading of imprecise diagrams. Technical Report 2006/16; Department of Computing, The Open University.

DOI: https://doi.org/10.21954/ou.ro.00016046

Abstract

In this paper we describe our approach to the grading (marking) of graph-based diagrams for instance those produced by students in the area of Entity-Relationship (E-R) diagrams. This is an application of a more general problem and is based on our framework for diagram understanding. We believe thattechniques of NLP commonly used in understanding text have equivalents in the understanding of diagrams and this paper shows how these techniques have been incorporated into a grading tool. The accuracy of the marking tool has been measured in two small-scale trials and the positive results from those trials are presented here. Our approach naturally allows the provision of diagrammatic feedback on student answers to E-R diagramming questions which has been incorporated into a tool for practising E-R diagramming.