Cloud computing and mapreduce for reliability and scalability of ubiquitous learning systems

Abstract

Ubiquitous learning research is about seamlessly enabling learning through the use of sensors that gather dataw from the learner surroundings and adapt learning contents accordingly. Nowadays, mobile devices play a big part of these systems due to their advanced capabilities, like communicating with other devices through different methods and standards, and the ability to connect other gadgets expanding functionality even further. On the other hand, sensors integrated with these systems are very advanced and sophisticated making it possible to gather tremendous amount of data. In this paper a new Ubiquitous Learning system architecture is presented. This new architecture enabled solutions for some major challenges in the field. A Ubiquitous Learning system design and implementation is presented as a use case. The system adapts learning contents based on applying an understanding degree detection algorithm on an input of brain signals collected from the learning student. The adapted learning contents are then sent back over to be displayed on a mobile device. The main focus of this paper is to show how the new architecture supports the necessary reliability and scalability for such systems. I'm proposing in this paper that using Cloud Computing and MapReduce as the architecture main building blocks lead to better approaches and solutions for these two problems. Evaluation showed the effectiveness of using the proposed architecture to support systems with an increasing number of users.