A graph-based method for ranking of cloud service providers

With the advent of cloud computing, more businesses are turning to the cloud for deploying their applications and for infrastructure solutions. Quality of Service parameters has a direct impact on businesses. Enterprises have to select the best cloud service provider with optimum cost. In this paper, we present a graph-based method for ranking cloud service providers. First, we compute a partial correlation between cloud service providers in terms of their response time. Second, we construct a graphical lasso regularization network with a penalty, which controls spurious connections. Third, the service providers are ranked based on degree centrality. Finally, we applied a normalized discounted cumulative gain method to measure the rank quality of cloud service providers. The comparative experimental results show that lasso regularization performs better than the traditional Bonferroni correction method.