A Tool for Behaviour-Based Discovery of Approximately Matching Web Services

We present a tool that is able to discover stateful Web Services in a database, and to rank the results according to a similarity score expressing the affinities between each of them and a user-submitted query. To determine these affinities, we take behaviour into account, both of the user’s query and of the services. The names of service operations, their order of invocation, and their parameters may differ from those required by the actual user, which necessitates using similarity scores, and hence the notion of soft constraints. The final tool is based on Soft Constraint Automata and an approximate bisimulation among them, modeled and solved as a Constraint Optimisation Problem.