Architectural backpropagation support for managing ambiguous context in smart environments

Abstract

The evolution to ubiquitous information and communication networks is evident. Technology is emerging that connects everyday objects and embeds intelligence in our environment. In the Internet of Things, smart objects collect context information from various sources to turn a static environment into a smart and proactive one. Managing the ambiguous nature of context information will be crucial to select relevant information for the tasks at hand. In this paper we present a vector space model that uses context quality parameters to manage context ambiguity and to identity irrelevant context providers. We also discuss backpropagation applied in the network architecture to filter unused context information in the network as close to the source as possible. Experiments show that our contribution not only reduces the amount of useless information a smart object deals with, but also the distribution of unused context information throughout the network architecture. © Springer-Verlag Berlin Heidelberg 2007.