An adaptive framework to provide personalisation for mobile learners

Abstract

The advancement of technologies in wireless and hand-held devices, combined with the ability to access learning content everywhere and anytime, has created signi cant interest in mobile learning (m-learning) in recent years. New smart phones are capable of exchanging voice, text, pictures and video. In addition, the new wireless network provides high-speed connections at a low cost to mobile users. Mobile learning ful ls the promise of learning "on the move" by allowing learners to take control over the time and location of their learning. Learning through a mobile device makes learning truly personalised. Learners have the ability to choose learning content based on their interests, thus making learning learner-centric. In contrast to typical electronic learning (e-learning) products, this access to personalised information means each learner can access the resources they need in a timely manner while minimising wasted bandwidth. Providing immediate access to relevant and interesting information, based on the individual learner's requirements, encourages use and increases engagement because learners are able to access the information they want wherever they are. Mobile learning is still in its infancy and research indicates that few projects have produced any lasting outcomes. Other research to date has focused on the connectivity problem of using wireless networks. The ultimate goal of this thesis is to present the design and implementation of a Machine Learning Based Framework for Adaptive Mobile Learning, to provide a logical structure for the process of adapting learning content to satisfy individual learner characteristics by taking into consideration the learner's needs. One main contribution of this thesis is a novel development framework in the eld of mobile learning. The framework depicts the process of adapting learning content to satisfy individual learner characteristics by taking into consideration the learner's needs. The system architecture of the context adaptation based learner pro le framework is fundamentally grounded on a number of logical layers. This framework provides a way to reduce the complexity of managing di erent mobile device settings to enhance learning environments. Delivery options for mobile learning are increasing, however new technologies alone will not improve the experience of mobile learners. There are a number of factors that impact on a typical learning experience, and many more when that learning experience becomes 'mobile'. This thesis presents an Adaptive Mobile Learning Content Framework that describes the factors that play an important role in delivering learning content to mobile learners, and their relationship with each other. Once the necessary information is collected about a learner - either automatically (e.g. location,device, previous usage) or through learner input (e.g. age) - learning content can be adapted to meet the unique and personal needs of that learner within their current context. It allows consideration of individual learning styles and scenarios, device and application capabilities, and material structure, leading to a customisation of the type and delivery format of learning information in response to the learner. Based on the newly developed frameworks, another major contribution of the thesis is the establishment of an adaptive learner model. Generally speaking, a m-learning adaptive framework provide personalised services to learners in accordance with their current situation or assumptions about each interacting learner. Adaptive systems adapt their own behaviour to suit and nd the optimal outcome for a speci c learner's needs. An effcient adaptive system is capable of deciding autonomously what to deliver, how to do it and when to do it. It is essential for adaptive systems to gather information about the learner. Without such information about the learner, the adaptive system is not able to adapt itself to the learner's characteristics and preferences. The required information is stored and managed in the form of a learner pro le and learner model. The construction of the learner model is another main contribution of this thesis. The vast amount of data involved in any successful adaptation process creates complexity and poses serious challenges. The Enhanced Learner Model focuses on how to model the learner and all possible contexts in an extensible way that can be used for personalisation in mobile learning. The learner model is logically partitioned into smaller elements or classes in the form of a learner pro le, which can represent the entire learning process. Learner pro- le contains learner's preferences, knowledge, goals, plans, place and possibly other relevant aspects that are used to provide personalised learning content. This thesis presents a Neuro-Fuzzy model for delivering adapted learning content to mobile learners. The adaptation of learning content is based on Adaptive Neuro-Fuzzy Inference System (ANFIS). ANFIS has been recognized for its exible and adaptive characteristics. ANFIS is a powerful approach to develop fuzzy systems that are capable of learning by providing IF-THEN fuzzy rules in linguistic form. The ANFIS approach is adopted to determine all possible conditions; these cannot be determined by using individual techniquea. The detailed simulation results demonstrated that ANFIS would help the adaptive system to determine a suitable learning content format.