Building Learner Profile in Adaptive e-Learning Systems

Researches and applications of e-Learning start with building course contents and materials based on Internet, besides the development of Learning Management Systems. However, in most early e-Learning systems, learners are often provided with identical services and learning materials, in the form of "one size fits all". That means all learners are the same; the systems do not discriminate the learner's backgrounds, learning goals and personal interests. Now, profile is being studied to apply widely in the e-Learning systems, in which the adaptation focuses on the goal of how helping learner to get knowledge, take interest in the learning activities effectively and suitably fit to learner. The core idea of our paper is to suggest a generic user model that supports more teaching activities and provides appropriately learning resources and services to each learner in the blended-learning environment: traditional learning in the classroom and on-line training. In this article, we will discuss building learner profile structure, especially initiating and updating profile. Structure of profile includes information components about demographics, training experiences, self-study activities and learning demands, in which each component can have the needed features to represent completely individual characteristics of learner when he/she joins in any training course of system. It is based on psychological and pedagogical foundations that we will present completely and clearly in article content. Besides, we deal with some main problems in the "classical" initiating and updating process based on explicit and implicit feedbacks. The updating process that is only based on the feedbacks still has challenges even when the correctness of feedbacks is ensured. Thus, we use rule-based induction approach for initiating and updating process. The inductive method is based on the alpha-Community Spaces Model. The basic idea of this approach is the feature values of a learner's profile that can be inferred basing on the profile of members in the same communities or learning groups with him/her. We have offered the architecture of Adaptive e-Learning System together with the proposed user model which has both the adaptability and communication support among the learners in learning activities formed from profile features. From the view of a personalized system, it aims at improving interactions between learner and materials, learner and instructor, and learners in group. Finally, some illustrated examples concretely for taking advantage of learner profile in system are also presented.