earticle

영어

In ubiquitous computing environments, users usually compose several services to achieve their goals. The service composition method of this paper is based on WSUN (Web Services on Universal Networks) [7]. It supports dynamic service composition which takes account of the dynamic factors of ubiquitous environments. In service composition, an operation of a service should be substituted by another operation when the existing operation is not available. The substitutable operation should support same functionality of existing operation. Moreover, it should be able to communicate with operation back and forth. The service composition mechanism of WSUN considers the functional similarity between unavailable operations and substitutable operation, but it does not take account of input and output message structures and parameter types of operations. In order to enhance WSUN service substitution method, we use SAWSDL (Semantic Annotation WSDL for XML Schema) [15], which is Semantic Web Services technique. SAWSDL provides two construct: modelReference and schemaMapping. A modelReference specifies the association between a WSDL or XML Schema component and a concept in some semantic data. Semantic annotations using modelReference help to discover substitutable operations. In addition, a schemaMapping solves the heterogeneity of I/O message structures and parameter types of operations. Consequently, even if the operation, which has same functionality of unavailable operation, has different I/O message structures and parameter types of unavailable operation. In order to recommend the operation has good quality, we define the quality of operation. It is a criteria that user can choose the best operation. In addition, we take a user weight for accommodation of user’s requirements. The recommendation system prioritizes operation through calculating quality of operation based on user weight value. Therefore, operation by operation’s priority recommends users. Our system supports substitutable operation discovery mechanism using Semantic Web Services technique, and substitutable operations list to satisfy user’s requirement recommend in ubiquitous environment. Therefore, user can discover substitutable operation in substitution situation. In addition, substitutable operations to fit user requirement recommend user according to weight value from user.