A geo-location context-aware mobile learning system with adaptive correlation computing methods

This paper proposes a context-aware mobile learning system with adaptive correlation computing methods. This system enables users to enhance their knowledge by correlating it with daily experiences. The proposed system contains a hybrid metric vector space to define the correlation between heterogeneous metadata vectors of the user context and learning material. The system integrates heterogeneous metric vector spaces with definitions of the semantic relations between the vector spaces. The significant feature of this system is a hybrid adaptation mechanism for the calculation of correlation. The adaptation mechanism has multidirectional adaptation functions for various learning materials, situations, and learners. We propose a revise-localize-personalize (RLP) adaptation model. In the adaptation mechanism, users only have to improve the metadata or the relations just in their relevant field. The advantage of the system is that the system reduces the time-intensive efforts required for describing direct relations between user contexts and learning materials. This paper presents the feasibility of the context-aware heterogeneous information provision with the hybrid metric vector space, by implementing an actual mobile application system and examining real-world experiments on data provision.