A Vagueness Adaptive Efficiency Evaluation Method of Science and Technology Resources Opening and Sharing Platforms Based on FAHP and BP Neural Network
针对科技资源开放共享平台效率评价中主观信息模糊的问题，融合模糊层次分析法和BP神经网络，构建了考虑服务数量、质量和效果的评价模型，并用中国六个国家级平台验证了其准确性和可靠性。
With the vigorous development of big data and information technology, promoting science and technology resources opening and sharing (STROS) has become a promising strategy to develop national innovation capacity. China has established various STROS platforms (STROSP) at the national and regional levels to encourage resources and knowledge sharing. However, STROSP efficiency evaluation is challenged in a vague environment, in which subjective and imprecise information in acquiring evaluation preferences of decision makers is a key obstacle. To solve these problems, an innovative model for STROSP efficiency evaluation is developed. An indicator assessment system explicitly considers the key evaluation aspects, namely, service quantity, service quality, and service effect. This article combines fuzzy analytic hierarchy process (FAHP) and backpropagation (BP) neural network algorithm into an integrated model to quantify the efficiency of STROS. Prior knowledge of experts was fully utilized by FAHP. The neural network algorithm enabled the intelligent extraction and rapid inference of sample data features. The proposed model maximizes fuzzy mathematics in solving fuzzy and nonquantifiable problems and utilizes the advantages of the BP neural network on nonlinear mapping. The accuracy and reliability of the model are validated by a case study of six national STROSP in China. Estimation results demonstrate that the model is a powerful method for the real-time evaluation of STROS efficiency evaluation.