Preference Model Driven Services Selection

Service, as a computing and business paradigm, is gaining daily growing attention, which is being recognized and adopted by more and more people. For all involved players, it is inevitable to face service selection situations where multiple qualities of services criteria needs to be taken into account, and complex interrelationships between different impact factors and actors need to be understood and traded off. In this paper, we propose using goal and agent-based preference models, represented with annotated NFR/i* framework to drive these decision making activities. Particularly, we present how we enhance the modeling language with quantitative preference information based on input from domain experts and end users, how softgoals interrelationships graph can be used to group impact factors with common focus, and how actor dependency models can be used to represent and evaluate alternative services decisions. We illustrate the proposed approach with an example scenario of provider selection for logistics.