The Research Of Semantic Web Services Discovery Of Spatial Information

Web services as an emerging distributed computing model, to be quickly applied to the field of spatial information because of it provides an effective mechanism of information integration and interoperability on the Web. Spatial Information Services combines Web services technology and geographic information systems (GIS), can provided all kinds of spatial information processing, which makes it possible of information acquisition, processing, sharing, integration and interoperability. With the development of new industries and the growth in demand for spatial information, major international companies and organizations have launched the standards and norms of Spatial Information Services. The OGC (Open GIS Consortium, an open GIS Association) proposed OWS (OpenGIS Web Services) as Spatial Information Services interoperability standard recommendation , ISO/TC211 (Geographic Information Science and the International Organization for Standardization Technical Committee) introduced the ISO 19119 Normative of Spatial Information Services. However, the existing industry-standard Web services as well as the norms launched by international organizations, can not express the semantics of Spatial Information Services and resolve the corresponding semantic conflict. It is difficult to meet users of Spatial Information Services applications in the Web environment. As spatial information play a more important role in many areas, Spatial Information Services needs to enhance the semantics and improve the interaction of semantic expression, and meet the urgent requirements. To find new Spatial Information Services model becomes a key issue of spatial information interoperation. In view of this, the paper makes in-depth study of Spatial Information Services description standard and Spatial Information Services discovery method.Semantic web technology can solve spatial information's semantic heterogeneity and realize the spatial information sharing and semantic interoperability. Because WSDL do not have the semantic information, W3C launched a special semantic service ontology language OWL-S, which provides a computer understandable formalization to the characteristics and capabilities of Web services. In this paper, through a combination of OWL-S, we do some work in Spatial Information Services description and discovery. The work summary as follows:First, for Spatial Information Services description, we present a new description based on Description Logics (DLs). Then we clarify this formal description through examples and discuss the reasoning capability in service discovery. DLs are a family of semantic knowledge representation languages, which has a good level of concept hierarchy and supports logic reasoning. Spatial information knowledge formalization based on DLs can accurately capture the field of geospatial information and make a common understanding of knowledge in geospatial field. In addition, we use first order logic (FOL) to express the precondition and effect as an extension of the function information of Spatial Information Services, in order to improve the efficiency of service query.Second, for Spatial Information Services discovery, the traditional UDDI and OGC Catalog are based on keywords searching, which do not support semantic constraints on fuzzy matching, and the semantic conflict has not been resolved. In order to improve the recall and precision, we research on the key issues in service discovery– Service Matching algorithm. Based on the flexible matching algorithm of Massimo Paolucci, we present a new algorithm of Spatial Information Services discovery - SSC algorithm.SSC algorithm improved in the following aspects:1. Introduction of Syntax Matching to match the non-functional information, such as the ServiceName and ServiceTextDescription, which adopts Edit Distance algorithm and LCS algorithm.2. Improvement of Semantic Matching. In flexible matching algorithm, the matching result which based on the ontology class inheritance relationship will be divided into four categories: Exact, subsumes, plugin, fail. This division particle is too large, which causes too many services appropriate was filtered by MatchMaker. In this paper we import semantic similarity between ontology concepts instead of the four categories. Through an example of outputs matches, we clarify the similarity calculation method and formula in the same ontology. Both the path length and the information capacity of the concept node take into account in this method. Also, we introduce WordNet to the calculation of concept similarity between ontologies.3. Introduction of the precondition matching algorithm based on first order logic.4. Addition of Customer Weight to express the degree of user's affirmation of spatial information.Finally, we designed and realized a service discovery engine MatchMaker prototype– GeoDisTool (SSC algorithm). We import OWL-S MX MatchMaker (developed by Matthias Klusch - Senior Researcher of German Research Center for Artificial Intelligence) for algorithm performance comparison. The main algorithm of OWL-S MX MatchMaker used Massimo Paolucci flexible matching algorithm. Through the analysis of recall, precision, Top-k precision of test results, we can conclude that SSC algorithm do improved the performance and efficiency of Spatial Information Services discovery.In a word, the study results of this paper have both theoretical significance and practical value,it can be applied to the semantic expression of Spatial Information Services as well as service retrieval. And there is still a lot of further research to be done in the future.