传统网络SBM效率评价方法在选择参考集时忽略了有效单元组成的"有效面集",造成评价结果偏低,无法反映低效生产单元改进为有效的最优路径。在可变规模报酬的假设下,选择生产系统的最大有效面集为参考集,构建了内部子过程存在关联的多阶段生产系统的基于最大有效面集的网络SBM效率评价模型,模型中子过程的权重设定充分考虑投入、中间产品和产出在子过程的价值分布。该模型实际反映了低效生产单元改进为有效的最优路径,具有较强的实用性。此外,模型的改进思路具有普适性,既可以在更为复杂的生产系统的网络SBM效率评价模型中进行延伸,也可以扩展到众多其它类型的网络DEA效率模型的构建中。最后,采用江苏省44家国家级科技企业孵化器的经营过程为实例来验证改进模型的有效性和优势。
The traditional network SBM efficiency evaluation method ignores the "frontier set" formed by the effective DUMs when the reference set is selected, leading to the lower evaluation result which can't reflect the optimal improved path of the inefficient DUMs. Under the hypothesis of variable scale reward, we choose the maximum frontier set of the production system as reference set, build the network SBM based on maximum frontier set to the multi-stage production system which has correlated internal sub-processes.The model fully considers the value distribution of inputs, intermediate products and outputs when set up weights of sub-processes. This model reflects the optimal path of inefficient DUMs to improve its efficiency actually, with strong practicability. In addition, the modified thought of the model is universal applicable. It can extend not only to network SBM efficiency evaluation model of the more complex system, but also to building the other types of network DEA efficiency evaluation model. Finally, this paper uses business process of 44 state- level science and technology business incubators in Jiangsu as an example to verify the effectiveness and advantage of the improved model.