A model for composite service discovery based on data dependency

Service discovery is the premise of service composition. The existing service discovery methods only consider individual service functions and static properties, less considering the inherent dependencies between the composite services. The data dependencies between services reveal the logic correlation between the composite services, but it is of great significance for the service composition and service discovery. Propose a domain ontology services dependency graph (DOSDG) model of composite service discovery based on the data dependence relations. This model consists of two layers composed of the lower from the domain services dependency graph (DSDG) and the higher domain ontology agent (DOA). Each DOA manages a DSDG, and all of the DOA forms a P2P structure. Reasoning the service domain context and the correlation between the services from user´s composite request can improve the efficiency in runtime. This approach takes into account the service domain features and the correlation between services, so it not only greatly reduces the services search space and can amend the user´s request.