Graph-Based Intelligent Ranking Model for Cloud Service Provider Selection

Abstract

The fast pace of the development of cloud computing has led to a relatively large number of cloud service providers (CSPs), and the choice of the most appropriate CSP has become a complicated problem of decision-making. The commonly used methods of cloud service selection usually employ standard multi-criteria decision-making methods, which do not reflect the intricate connections between the service attributes, user needs, and the performance of providers. Within the framework proposed, cloud vendors, service characteristics, quality-of-service (QoS) parameters, and preferences of users are represented as interrelated graph entities. Graph based analysis is used to determine the hidden relations and dependencies between various selection criteria and an intelligent ranking mechanism ranks providers according to performance, reliability and security, cost efficiency, scaling and availability of services. The model will address the shortcomings of independent attribute-based ranking since it takes into account the structural relationships within the cloud service ecosystem. Experimental analysis of the data on cloud services shows that the proposed method is effective to rank the appropriate providers, simplify the selection process, and increase the efficiency of the decision-making process in contrast to traditional ranking tools. The intelligent cloud service selection model proposed is a scalable and adaptive method of intelligent selection of cloud services in dynamic cloud systems.