Contents I. Introduction
It is a trend that dynamic composition in service-oriented architecture enables the runtime discovery and composition of published services and data without altering the client side applications[1]. It is very important for enterprises to utilize disparate services and available data on web for integrating a system quickly and with little cost. This is dynamic composition of services and data. There are many research works about this problem. Fujii presents a semantic-based dynamic service composition system[2] which provides a functionality of semantics-awareness, middleware, and semantics-based service composition mechanism. Penta describes WS binder which is a framework for making dynamic binding of service compositions feasible according to functional and nonfunctional constraints[3]. It supports three kinds of binding types including pre-execution binding, run-time binding and run-time rebinding. Laliwala proposes a semantic-based and rule-based dynamic composition method which generates business processes[4]. The generation of business processes is based on event-condition-action rule, forward chain algorithms and backward chain algorithms. Chafle proposes a solution for creating and executing web services, which can adapt to changes in the operating environment at the stages of composition, deployment, and runtime[5]. Enterprise service bus provides a highly distributed and event-driven environment for service oriented architecture. ESB is often used as the infrastructure for service connection and message exchange in SOA[6] [7]. A key feature of ESB-enabled SOA is that the service definition is separated from locating and invoking services.
