A novel method for learner assessment based on learner annotations

Abstract

Assessment is one of the most essential parts of any instructive learning process which aims to evaluate a learner's knowledge about learning concepts. In this work, a new method for learner assessment based on learner annotations is presented. The proposed method exploits the M-BLEU algorithm to find the most similar reference annotations and then the learner annotation will be processed further to check essential words, words order and contradictions. To examine this new approach, a virtual learning environment was designed and implemented in which assessment of the learner's knowledge is performed on the basis of main and sub concepts. These concepts are delivered by means of course contents and the learning environment guides the user to annotate concepts. Evaluation results show that our designed system can effectively assess learner's knowledge. The benefit of suggested assessment method is its implicitness of assessment approach. Furthermore the correct annotations can be used to help the users remembering concepts by reviewing their annotations. © International Forum of Educational Technology & Society (IFETS).