Recommender Model for Messaging System at MOODLE

Abstract

E-learning systems are powerful way for enriching learning and teaching process. With the spread of this culture, there is a need for Learning management system as MOODLE, aTutor, canvas and Sakai. Modular Object-Oriented Dynamic Learning Environment (MOODLE) allows students and instructors to interact easily and effectively in various ways, especially via private messaging system. In order to make this feature more convenient for both students and instructors, this research proposed a model that integrated within messaging system of MOODLE to filter unwanted messages, recommend a material for students, and to help them finding answers to their questions which are founded on the message. The problem of this research is how to build a model to prevent message which contains taboo words and send recommended material to students according to their questions. This new model will make it easier for students to find information and materials they are looking for, or asking about and at the same time it will keep instructors' mailbox clean of any unwanted irrelevant messages. The research methodology consists of four stages start from stating the theoretical foundation, then designing the model functions, building the model's functions, and finally evaluating the Model. The model was applied at University of Palestine on six courses with various natural of materials, some of these materials are isolated topics. In this type of courses the accuracy of the system was high. On the other hand, there are some courses in which their topics are integrated with each other, and the accuracy was less than the other type of materials, but in general the whole system's accuracy is 96.66% and the reliability of the system is 92.55%. Finally, the model is founded to improve using message in the MOODLE, and the model actually solved the problem of this research.