چكيده به لاتين
nowadays, competition in organizations has reached a high level and if organizations want to continue to work in this competitive environment, they need to be aware of their situation. one way to determine the status of the organization is to evaluate the performance of its internal units. The holding or parent company has a headquarter and a queue. The purpose of this study is to evaluate the performance of Iran's maritime headquarter holding units. Data envelopment analysis is one of the effective methods in estimating relative efficiency. In this research, a network data envelopment analysis approach has been utilized to make the organization's processes better, more accurate and more rational given the organization's reality. Accordingly, 11 maritime headquarter units were evaluated. After defining the indicators for performance evaluation using the opinions of academic and organization experts, a two-stage network DEA model was applied which the data related to the indicators were collected from various sources. Then, the pairwise comparisons matrix was used to determine the importance of each unit in the indices and the raw data were adjusted. Finally, using two DEA models without network and two network models of headquarters in both raw (unmodified) and modified data modes, were evaluated and ranked. It is notable to say that the gams optimization software was used to solve the mathematical model. The results indicate that in the assessment of the entire network, headquarter units of planning and programming, and quality have an efficiency of 100 % and jointly ranked first, followed by the human resource management unit with 77% efficiency. Also, the weakest performance was assigned to the support headquarters and ranked last. It is notable that data modification has had a significant impact on the efficiency and ranking of the units. For example, the engineering unit that ranked first in all models before adjusting data fell to lower ranks after that. Also, some units such as planning and programming, which have low efficiency before data adjustment, have achieved high performance after data modification. The influence of the intermediate processes in the overall performance of the network is extremely important. The proposed model has the advantage over other performance evaluation models as well as traditional non-network models of data envelopment analysis to estimate the efficiency of multi-stage processes and network structures.
