Ontology Based Service Discovery Method for Internet of Things

Abstract

As the development of internet of things (IOT), massive sensors have been deployed as the public infrastructure. However, there is a bottleneck in state of the art IOT applications, which require sensors to response on abstract and complex requests similar to human natural languages. On the contrary, since sensors' service directories are unable to map their services on how users' abstract queries can be performed, sensors can only response to low-level queries with accurate and complete descriptions. To handle this challenge, in this paper, we proposed a novel ontology based method to improve sensor systems' reasoning. The key lies on any abstract terminologies can be understood by searching their ontologies descriptions. Therefore, by utilizing ontologies, abstract queries can be hierarchically decomposed into several simple queries handled by single sensors. Moreover, ontologies also imply the logical relations on the feedbacks from each single queries, and the answer can be carried out based on the equations and sensor outputs.