Integration of sensor networks and linked data in a coastal flooding scenario

One of the issues of current interest in sensor networks is how to discover and integrate data from a wide range of sources: from data persisted in legacy information systems or coming from sensor networks to data available in the Web (i.e., linked data). One way to facilitate disparate data discovery and integration is to describe them semantically. The SemsorGrid4Env project is working on systems where data sources can be located with semantic web technologies while defacto standards can be used to actually access the data. By building a demonstration around a coastal flooding scenario we had to build bridges between the terms used by domain experts and those used by data providers and to develop the necessary data integration mechanisms supported by this semantic reconciliation. This paper will describe the issues involved in putting together the system and will illustrate how a linked-data approach can be used.