Preference-Based Discovery of Dynamically Generated Service Offers

The majority of service discovery approaches operate on abstract service descriptions. However, a single service often provides a significant number of possible service offers which are not reflected in abstract service descriptions. In this paper, we define a preference-based discovery model which operates on rich search request descriptions and dynamically generated individual service offers. We define search request model that include hard constraints, rich preferences, and flexible input parameters. We use a combination of utility functions and weighted rules for modeling rich preferences. We apply our results to an international shipping scenario in the experiment to prove the feasibility and usefulness of our approach in a realistic scenario.