Automated Evaluation of Learners with ODALA: Application to Relational Databases E-learning

Abstract

This paper deals with an approach for an automated evaluation of the learners' state of knowledge when learning by doing. This approach is called ODALA for "Ontology-Driven Auto-evaluation for e-Learning Approach". It takes place in the context of Computer Based Human Learning Environment (CBHLE) in a self-learning by doing mode. ODALA is based on the teaching domain ontology and on errors classification and detection. The evaluation process is composed of four stages: (1) form analysis of learner's solutions, (2) semantic analysis, (3) marking, and (4) updating of the learner's model. We bring the approach into play in the context of relational databases teaching: we present the results of the relational databases self-learning system (RDB-E-LEARN) development, where the main stages of our evaluation approach are implemented.

Keywords

• Self-learning
• Error diagnosis
• Learner's evaluation
• Learner's marking
• Learner's model
• E-learning.