Wearable context recognition for distance learning systems: A case study in extracting user interests

Abstract

We propose a novel approach to context recognition for use in extracting user interests in e-learning systems. Under our approach, a series of screen images are captured by an imaging device worn by the user engaged in elearning. From these images, vendor logo information is detected and used to deduce the e-learning context. The compiled history of recognized context and elearning access can then be compared to extract low-interest topics. Experimental results show that the proposed approach is robust and able to identify the proper context 96% of the time. © Springer International Publishing Switzerland 2013.