Educational multimedia adaptation for power-saving in mobile learning

Abstract

Limited-battery power is a major constraint in mobile learning. It is important to adopt battery power-saving mechanisms in mobile learning applications in order to extend the duration of learning activities. This thesis explores issues related to power-saving in mobile learning. Streaming of online educational multimedia on mobile devices is a power hungry activity due to large amount of wireless data transfer. A number of power-saving multimedia adaptation techniques for streaming multimedia have been developed in the past. Most of these existing approaches achieve power-saving by uniformly lowering the presentation quality of an entire multimedia stream. These generic techniques will typically lower the visual quality of an entire multimedia stream uniformly, without considering its impact on perceived loss of visual information at different points of the multimedia stream.

In this thesis, through a user study we suggest that reducing the quality of educational multimedia beyond a certain level - for power-saving adaptation - can cause perceived loss of visual information in quality-sensitive portions of a multimedia. This could have a negative impact on perceived learning effects and leave the resource unsuitable for learning. The results of the study suggest that different parts of a learning multimedia may have different lowest acceptable presentation quality requirements for avoiding perceived loss of visual information. The participants of the study were able to comprehend visual information in one fragment at a lower visual quality but could not comprehend visual information of some other fragments at the same quality level. To address this problem, we proposed a Content-Aware Power Saving Educational Multimedia Adaptation (CAPS-EMA) approach that suggests a way of delivering each portion of a multimedia in a lowest acceptable quality based on the visual contents of each fragment. We demonstrate an implementation of this approach using a prototype system called MoBELearn. The results of our evaluation studies suggest that the way CAPS-EMA adapts multimedia resources is acceptable to users in power-saving situations. CAPS-EMA requires some authoring processes in order to identify fragments and lowest acceptable quality constraints. An expert evaluation described the activities involved in the authoring process as easy to understand and perform

Power-saving multimedia adaptation mostly results in some compromises in terms of visual quality and information content. Existing techniques offer users little control over the adaptation process and they are obliged to accept the consequences of the adaptation. We propose a Learner Battery Interaction (LBI) mechanism that suggests offering users power-saving options and relevant feedback about the expected compromises for each power-saving option. This would enable users to make informed choices about power-saving. We evaluated the concept of LBI through a user study. The results of the study suggest a positive perceived usefulness of the system and that mobile learning applications may benefit from the idea.

In the end, we propose a search mechanism for online adaptive learning resources that would help find a personalised learning resource that would fulfil the information needs of a learner in a battery-efficient way. This proposed mechanism is based on the concept of discovery of online open adaptive learning resources. For this purpose, we proposed an ontology model to describe adaptive learning resources, in terms of its adaptive features: learning and presentation features. This model could be used as a basis for implementing the proposed concept of the discovery of versions of adaptive learning resources in order to enable learners to engage in learning activities in a battery-efficient way by searching for online learning resources.

More information

Identifiers

ORCID for Nicholas Gibbins: orcid.org/0000-0002-6140-9956

Catalogue record

Export record