On the Efficiency of Cloud Providers: A DEA Approach Incorporating Categorical Variables

Abstract

Cloud computing is a growing industry and it has already dominated many IT markets segments. Cloud providers offer numerous equivalent IaaS services aiming to fulfill clients' requirements. In addition, cloud clients need to choose solutions that minimize costs without compromising efficiency though. However, not only the confusion due to the large variety of cloud services but also the uncertainty about the efficiency specifications that their cloud services should have, can make cloud computing services selection a difficult task for the users. Into this context, this paper presents an approach to a multi-attribute decision-making problem that focuses on the calculation of efficiency of IaaS cloud services as a measurable driver for both clients and providers. A DEA input-oriented model is described, which estimates efficiency of cloud services based on functional and non-functional parameters. Furthermore, the contribution of functional and non-functional features to the overall performance is examined. This innovative model urges providers to optimize the efficiency of their services aiming to increase their market share and, at the same time, assists clients in choosing a cost-effective cloud solution.