LIVESGEO : a location-based multimedia knowledge sharing system

• Main Author: Chung, Yongik
• Language: English
• Published: University of British Columbia 2011
• Online Access: http://hdl.handle.net/2429/36762

Learning is vitally important because we live and work in a rapidly changing world. As learning is acquisition of knowledge or skills through utilizing given information, it is critical that one needs to be literate in order to learn effectively. However, conventional learning materials are mostly targeted for literate audiences and delivered by the means of written materials in classical educational settings making it harder for illiterate audiences to engage in learning. This thesis introduces a novel location-based multimedia knowledge sharing system, named LIVESGEO, which takes advantages of advanced capabilities of smart-phones and mobile data networks to allow peer driven learning. The system allows illiterate audiences to easily create, share, search and rate knowledge contents on the spot. The system proposes a hybrid of client-server architecture and peer-to-peer architecture to efficiently circumvent the critical limitations and leverage the advantages of mobile devices. We also propose a novel mathematical framework to characterize relative popularity of a rated content based on its reputation and quality and justify the performance with empirical evaluations against popular rating frameworks. We also explore the applicability and usefulness of geo-tagged contents accumulated by LIVESGEO. To demonstrate this, we abstract the system into a spatial content cluster framework, which acts as an abstract layer between reverse-geocoded contents and applications to provide information about the contents that are clustered based on their geo-location, to allow unique applications to be built upon the system. In addition, we provide examples of such applications which utilize the spatial content cluster framework. In order to justify the use of reverse-geocoding to cluster contents geographically, we provide empirical analyses of reverse-geocoding by making novel use of two popular mathematical models in the field of cartography, namely the Haversine and Vincenty formulae. === Science, Faculty of === Computer Science, Department of === Graduate