Towards an automated system for short-answer assessment using ontology mapping

Abstract

A key concern for any e-assessment tool (computer assisted assessment) is its efficiency in assessing the learner's knowledge, skill set and ability. Multiple-choice questions are the most common means of assessment used in e-assessment systems, and are also successful. An efficient e-assessment system should use variety of question types including short- answers, essays etc. and modes of response to assess learner's performance. In this paper, we consider the task of assessing short-answer questions. Several researches have been performed on the evaluation and assessment of short-answer questions and many products are deployed to assess the same as part of e-learning systems. We propose an automated system for assessing short-answers using ontology mapping. We also compare our approach with some existing systems and give an overall evaluation of experiment results, which shows that our approach using ontology mapping gives an optimized result.