Dynamic Web Service Composition within a Service-Oriented Architecture

Increasing automation requires open, distributed, service-oriented systems capable of multicriteria-driven, dynamic adaptation for appropriate response to chang- ing operating conditions. We combine a simple architec- ture with a novel algorithm to enable openness, distribu- tion, and multi-criteria-driven service composition at run- time. The service-oriented architecture involves mediator web services coordinating other web services into composi- tions necessary to fulfil user requests. By basing mediator services' behavior on a novel multicriteria-driven (includ- ing quality of service, deadline, reputation, cost, and user preferences) reinforcement learning algorithm, which inte- grates the exploitation of acquired knowledge with optimal, undirected, continual exploration, we ensure that the system is responsive to changes in the availability of web services. The reported experiments indicate the algorithm behaves as expected and outperforms two standard approaches.