Towards a semantics extractor for interoperability of IoT platforms

Achieving the interoperability of Internet of Things (IoT) platforms raises a key challenge, as most current IoT applications are vertically integrated within separate silos without horizontal communications. In response to this, we propose the Semantics Extractor, a framework for automatic extraction of semantic graphs from REST interfaces exposed by IoT platforms. Through an iterative process of extracting, identifying, inferring, enriching and refining, our framework is able to discover IoT resources, add semantics to platforms' payloads and capture latent links within and between resources exposed by IoT platforms, from which further semantics can be deduced. We show via examples how RDF graphs are incrementally generated and enriched by such iteration process to interoperate IoT platforms. At last we present an implementation architecture.