Hoarding Content in an M-Learning System

The advances in mobile technologies predispose the support of learners’ and teachers’ activities on the move. As the nowadays’ infrastructure still does not provide cheap ‘always on’ connection at every place and every moment, a mobile learning system should provide also support for offline usage of learning material. The process of automatically selecting and pre-fetching content to be used during the disconnected periods is called hoarding. Considering a situation where the mobile device’s limited memory can not hold all available materials, the aim of the hoarding is to predict what the user will really need and use. A high accuracy of the set selection is necessary, as in the learning scenario every miss can lead to student’s misunderstanding of the studied subject. Hoarding is still not well explored. We are currently developing a mobile version of an existing language learning online platform. Our Mobile ELDIT system strongly needs a hoarding subsystem. We have developed a basic hoarding algorithm, whose preliminary outcomes we report and we want to study the parameters that could help a hoarding algorithm improvement in order to cover the peculiarity in m-learning scenario. Our goal is to provide an efficient strategy, taking into account additional parameters automatically extracted by the system.