Context-dependent Data-Envelopment-Analysis-Based Efffiiciency Evaluation of Coastal Ports in China Based on Social Network Analysis

An emerging trend in performance evaluation is combining social network analysis methods with data envelopment analysis (DEA) models and using network centrality methods to distinguish DEA results. One study employed an input oriented variable-returns-to-scale DEA model to address referent decision-making units and the corresponding lambda values to construct a network. This study referenced the literature and improved on the use of DEA weight sets to construct a network. We employed a context-dependent DEA model to delineate multiple effective frontier planes, aggregate the reference set relationships on each frontier plane to construct a network relationship matrix, and assess the influences of the interaction layers between the networks transformed by multiple frontier planes. Finally, our method was employed to evaluate the efficiency of coastal ports in China and rank ports by their efficiency. The results indicated that Qingdao Port was the most efficient, followed by Shenzhen Port; this finding verified the feasibility and rationality of the improved method. The present study contributes considerably to the theories on evaluation methods and identifying highly efficient ports.