EDA-based approach to comprehensive quality-aware automated semantic web service composition
Chen Wang, Hui Ma, Gang Chen · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2018
In the domain of Service-Oriented Architecture, web services are selected and composed to meet users' functional and non-functional requirements. A few researchers have proposed Evolutionary Computation (EC) techniques for service composition problems, where semantic matchmaking quality and Quality of Service (QoS) are individually or jointly optimised. However, these EC-based approaches allow best individuals (i.e., composition solutions) to be carried across generations, but does not consider the knowledge of the promising individuals for breeding better solutions. Due to its capability of learning the knowledge of good solutions, Estimation of Distribution Algorithm (EDA) has been successfully applied to tackle semi-automated QoS-aware service composition, where an abstract composition workflow is assumed known. However, this assumption is not always satisfied, so learning the knowledge of solutions without pre-given workflows is very challenge. We propose an EDA-based service composition approach to jointly optimize QoS and semantic matchmaking quality in a fully automated way and our experiment evaluation demonstrates the efficiency and the effectiveness of our proposed approach.