Research On The Semantic-based Distributed Data And Service Unified Discovery

With the rapid development of science and technology, Internet has become one of the necessities of daily life and a giant driving force of human society. Although network is a distributed information space containing great amount of knowledge with potential values, and people can have easy access to the information in the Internet, it is still difficult for them to get useful knowledge with ease. Current situation is that people don't lack of information, but they are heavily overwhelmed by it. Therefore, data and service discoveries are two most significant applications in the network, and almost all the network functions need their support.Current resource discovery systems almost deal with service discovery and data discovery separately, without considering the relationship between them. As a matter of fact, there are close connections between data and service. If we can exploit and make use of these connections, not only can we decrease the resource waste caused by the two separate discovery systems, but a convenient and universal discovery mechanism for users can also be provided. However, the research work and applications in this field are limited. Thus, based on the connections between data and service, this paper presents a unified description method for both of them, together with a semantic based distributed data and service unified discovery system. In addition, as classifying services with QoS (Quality of Service) is a development trend and significant feature of future networks, it is of great significance to provide service discovery with QoS for users. We propose to append QoS to OWL-S (Ontology Web Language for Services), and called it OWL-QoS. With the help of OWL-QoS, the description ontology of service includes QoS. Moreover, we also present a matching algorithm based on the OWL-QoS.The research work of this paper is carried out based on the national grand fundamental research 973 project, "Fundamental Research on the Architecture of Universal Trustworthy Network and Pervasive Services". It puts forward a new architecture of Network, which includes two layers, namely, Network Communication Layer and Pervasive Service Layer. Our research work is mainly applied in the Pervasive Service Layer, and has achieved two goals: a unified description of data and service, i.e., a generation of SID (Service Identifier), is proposed, and a mapping process from users' description to SID, which is equivalent to the network search engine, is also realized. The contribution of this paper is the main work of the Pervasive Service Layer, and it is also an important part of the national 973 project. The main contents and innovations are outlined as follows:(1) Aiming at data and service discovery, an ontology-based method for a unified description, via property connections between data and service, is proposed.(2) On the basis of OWL-S, the semantic service description with QoS and corresponding matching algorithms are proposed. Experimental results verify that the proposed service discovery approach with QoS is feasible and effective.(3) After evaluating the shortcomings of the keyword-based discovery method, a semantic similarity degree between concepts is presented, which is based upon the distance and granularity among the concepts. Moreover, a semantic based data and service unified discovery prototype system is initially proposed. Based on the prototype, a semantic unified matching approach of both data and service is presented, which realizes the universal registry and discovery of data and services. Experimental findings and performance analysis show that the proposed method can improve the search effectiveness, precision, and satisfaction degree of users.(4) The proposed approach for data and service unified discovery is applied to thenew network------the Universal Network, and the service identifier (SID) is generated.In addition, by combining data and service unified discovery method with the SID mapping mechanism of Universal Network, a mapping process from users' description to SID is established, and the data and service unified identification, registry and discovery of Universal Network are realized.