Enhancing the Discovery of Internet of Things-Based Data Services in Real-time Linked Dataspaces [chapter]

Wassim Derguech, Edward Curry, Sami Bhiri
2019 Real-time Linked Dataspaces  
A dataspace is an emerging data management approach used to tackle heterogeneous data integration in an incremental manner. Data sources that are participants in a dataspace can be of various types such as online services, open datasets, sensors, and smart devices. Given the dynamicity of dataspaces and the diversity of their data sources and user requirements, finding appropriate sources of data can be challenging for users. Thus, it is important to describe and organise data sources in the
more » ... aspace efficiently. In this chapter, we present an approach for organising and indexing data services based on their semantic descriptions and using a featureoriented model. We apply Formal Concept Analysis for organising and indexing the descriptions of sensor-based data services. We have experimented and validated the approach in a real-world smart environment which has been retrofitted with Internet of Things-based sensors observing energy, temperature, motion, and light. The chapter is structured as follows. Section 8.2 explores the need for discovery of data services within dataspaces. Section 8.3 provides an overview of existing semantic approaches to service discovery. Section 8.4 sets out the theoretical foundations of Formal Concept Analysis and shows how it can be applied to indexing Internet of Things-based data services based on their capabilities using a concept lattice. It then shows how the concept lattice can be used to discover sensors and relationships between sensor properties. Section 8.5 reports on a smart environment intelligent system validation of the discovery approach. Finally, Sect. 8.6 draws conclusions and details of future work.
doi:10.1007/978-3-030-29665-0_8 fatcat:wigxd3fvcbh5vmyutwle2c2ybi