Towards fuzzy QoS driven service selection with user requirements

Many QoS-aware service selection approaches assume that the QoS attributes are crisp values and the actual user requirements are not taken into consideration, when the service-oriented applications are constructed. As a result, users searching result may not be correct and good, because there are uncertainties in the data and the optimal solutions but not satisfying some requirements may not be acceptable to some users. In this paper, we propose to use Fuzzy Set Theory (FST) and fuzzy genetic algorithm (FGA) for QoS-based service selection. FST is applied to specify the triangular fuzzy-valued description of the QoS properties. A FGA is proposed to solve the QoS-aware service composition problem, which considers the actual QoS requirements from users in the selection process. Empirical comparisons with two algorithms on different scales of composite service indicate that FGA is highly competitive regards to searching capability.