Efficient distributed semantic based data and service unified discovery with one-dimensional semantic space

• Propose a scalable distributed system for data and service unified discovery.
• Adopt one-dimensional vector in semantic space to identify and locate both data and services.
• Specific services are represented by OWL-QoS.
• Improve the accuracy and efficiency of the searching results, and satisfy users much better.

Data and service discoveries are two significant applications in the Internet, and almost all the network functions need their support. A lot of research work has focused on either service discovery or data discovery respectively, although they cannot be separated. The existing network, due to its decentralized nature and weak support for semantic, is still chaotic and lacks the ability to allow users to discover, extract and integrate the information of interest from heterogeneous resources. This paper proposes a scalable, high performance distributed system for data and service unified discovery. In order to exclude ambiguity, our unified discovery system adopts one-dimensional vector in semantic space to generally identify and locate data and services, and then represent specific services by OWL Web Ontology Language for Services (OWL-S). Moreover, we employ Juxtapose (JXTA) architecture as a proper foundation to organize network peers. Experimental results illustrate that, compared with other discovery systems, the semantic-based unified discovery system can improve accuracy and efficiency of the search results, and satisfy users much better.