Pre-Evaluation with a Personalized Feedback Conversational Agent Integrated in Moodle

Abstract

Pre-evaluation of the learner's level is a common learning strategy designed to determine the prior knowledge and skills of learners. A pre-evaluation is carried out at the beginning of the course and based on the results obtained, personalized resources will be provided that respond to individual learner needs. This paper presents a pre-evaluation for a C programming language course by providing, at the end of the quiz, a personalized formative feedback and recommendation to the learners. We have developed our conversational chatbot named QuizCbot, which allows learners to go directly to the parts where they need the most help through the personalized feedback provided to them, including their final scores, the questions they answered correctly and the questions they answered incorrectly with the correct answer and explanation. Hence, the chatbot makes a recommendation on the concepts in which the learner did not obtain the average, identifying the concepts not mastered where the learner needs more (or less) support. Determining what learners know and don't know can help to improve the learning experience. We have integrated our QuizCbot chatbot, which is based on Natural Language Understanding (NLU), into the Moodle learning environment.