Enabling knowledge extraction from low level sensor data

Abstract

While sensor networks play a significant role in the modern information society, they output data in proprietary format and with little or no associated semantics. As a consequence, sensed data must be managed on a case by case basis, requiring significant human efforts. In this paper, we present an approach that: seamlessly supports any kind of network by exposing sensed data in a standard format; enables users to specify at a high level how to enrich sensed data with the semantics in which data is generate; facilitates end users in transforming data to meet their analytical requirements. © 2011 Springer-Verlag Berlin Heidelberg.