Abstract
Traditional learning systems work offline. Their output acquired after some unknown long time delay is a comprehensive instruction. However, the utility of such knowledge reduces in a time-constrained environment where instruction has to be completed by a certain deadline. As learning material scale up and longer instruction times become common, knowledge acquisition become time-dependent and is increasingly acquired on the fly due to considerable changes in today's know-how economy requirements. In this paper, we propose a procedural tradeoff between the time allocated to an instruction and the quality of the corresponding knowledge volume.