Data Services Match Based on Scene in Big Data

Existing web-services description is limited to the interface´s type, parameters, and operation´s definition and description, it is insufficient to describe the data-services in big data. Relative to the web-services, data-services is more fundamental, What´s more important is to describe the characteristics and semantic attribute of data source. In this paper, used Scalable OWL-S (Ontology Web Language for Services) to build data-services model, and proposed a method to solve the problem of low precision ratio and recall ratio in data-services match. Extracted unified data features and semantic description from data-services model using OWL-S, using a semi-supervised KNN-SVM (K-Nearest Neighbor-Support Vector Machine) classification method, classify the data-services based on scene and match data-services in the scene. Finally, the contrast experiment proved that the method is feasibility and effectiveness.