WEBLORS – a Personalized Web-Based Recommender System

Mohammad Belghis-Zadeh, Hazra Imran, Maiga Chang, Sabine Graf

Nowadays, personalization and adaptivity becomes more and more important in most systems. When it comes to education and learning, personalization can provide learners with better learning experiences by considering their needs and characteristics when presenting them with learning materials within courses in learning management systems (LMSs). However, the limited number of learning materials that exists in an LMS is not enough to provide students with different needs with rich personalized content. One way to provide students with more personal learning materials is to deliver personalized content from the web. However, due to information overload, finding relevant and personalized materials from the web remains a challenging task. This paper presents an adaptive recommender system called WEBLORS that uses a combination of web mining, text mining and recommendation techniques to discover, validate, rate and categorize relevant and personalized learning objects (LOs) from the web and present them to learners in an LMS. The system aims at helping learners to overcome the information overload by providing them with additional personalized learning materials from the web to increase their learning and performance. This paper also presents the evaluation of WEBLORS based on its recommender system acceptance using data from 36 participants. The evaluation showed that overall, participants had a positive experience interacting with WEBLORS. They trusted the recommendations and found them helpful to improve learning and performance, and they agreed that they would like to use the system again.