I-OnAR: a rule-based machine learning approach for intelligent assessment in an online learning environment

Abstract

Intelligent systems are created to automate decision making process that is similar to human intelligence. Incorporating intelligent component has achieved promising results in many applications, including in education. Intelligence modules in a tutoring system would bring the application and its capability closer to a human's ability to serve its human users and to solve problems. However, the majority of the online learning provided in the literature review especially in Malaysia, normally only provide the lecture notes, assignments and tests and rarely suggest or give feedbacks on what the students should study or do next in order to fully understand the subjects. Hence, the researchers propose an online learning environment called Intelligent Online Assessment and Revision (I-OnAR). It facilitates the learning process at multiple learning phases such as test creation, materials revision, feedback for improvement and performance analysis. These components are incorporated into the tutoring system to assist self-pace learning at anytime and anywhere. The intelligent agent uses a Rule-based Machine Learning method for the adaptive capabilities such as automated test creation and feedbacks for improvement. The system has been tested on a group of students and found to be useful to support learning process. The results have shown that 60% of the subjects’ performance have improved with the help of the system. The students were given feedbacks on the topic they did poorly as well as how to improve their performance. This proves that the Intelligent Online Assessment and revision (I-OnAR) can be a useful tool to help online students intelligently, systematically and efficiently. For the future works, the researchers would like to apply the usage of other techniques such as Fuzzy Logic to strengthen the analysis and decision of the current system.