Sensor networks play an important role in our modern information society. These networks are used for a variety of activities in different domains, including traffic monitoring, environmental analysis, transport and personal health. In general, systems generate data in their own format with little or no associated semantics. As a result, data must be managed individually and significant human effort is required to analyze data and develop ad-hoc applications for different end-user requirements. The research presented here proposes a holistic and comprehensive approach to significantly reduce the human effort in analyzing networks of sensors. The goal is to facilitate any form of sensor network, enabling users to combine related semantics with sensor data, and facilitate the end-user transformation of data necessary to provide more complex query expressions, and thus meet the analytical requirements.