Analyzing usage in a large coursecasting service
Owen G. McGrath · 2009
Coursecasting, the video and audio recording of lectures and events for live or on-demand playback, has become a significant IT service at many institutions of higher education. By making available on the internet collections of full-length digitized recordings, college and university information technology groups find themselves supporting large and varied distributed audiences--from local students to faraway elearners. Managing the up-to-date production and on-demand delivery of this kind of open content requires a technical infrastructure for capturing, publishing, and archiving that can adapt quickly to changing workload demands. Adapting quickly, moreover, requires special attention to understanding usage of the time-based content: the different kinds of users, where and how they access content, and shifting patterns of media access. Those who carry out usage analytics for open educational resource services involving internet-based video and audio face special challenges. This paper explores approaches to monitoring and analyzing audio and video content usage across two related but distinct dimensions-server workload and user behavior-in a large scale, open content coursecasting service.