Reactive Service Selection in Dynamic Service Environments

Due to the highly dynamic nature of services (web services can enter or leave the system at any time, or change their characteristics), adaptation to change during service composition is necessary to meet user needs. Yet current approaches to change handling detect quality violations and service unavailability only after their occurrence (after executing the corresponding service), resulting in undesired situations at execution time from which recovery (usually through costly replanning) might not always be possible. In response, this paper presents a novel reactive selection algorithm, which adapts to changes in the environment efficiently while performing the selection, ensuring that the selected composite service is executable, satisfactory and optimal prior to execution. The algorithm's effectiveness is demonstrated via experimental results.