Abstract:The evaluation of scientific and technological talents is a multi-criteria decision problem, and different methods lead to uncertainty in the evaluation, and the existing relevant evaluation methods have certain limitations in practice. To solve this problem, based on the idea of "breaking the four only" and "setting a new standard", in view of the differentiated requirements for evaluating scientific and technological talents at different stages of innovation activities, scientific and technological talents are classified into the categories of basic research, applied research and technological development、scientific and technological management、science and technology management services. Based on the keyword extraction method of policy documents and literature to construct the evaluation index system of scientific and technological talents classification, by fully considering the uncertainty of different methods leading to the existence of evaluation, combining the point allocation (PA) and the full consistency method (FUCOM) to construct the combination of grey number weights, and taking the weights of grey number and the evaluation scores of different decision makers for the evaluation object as the input of grey number, we propose the PA-FUCOM grey correlation analysis method; and empirically analyze four talents in the category of applied research and technology development from a research institute in Shaanxi Province as an example to ensure the validity and accuracy of the evaluation results. Grey system theory helps to solve the uncertainty in the evaluation process of scientific and technological talents, the point allocation is simple and feasible in practical evaluation, and FUCOM provides new opportunities for the calculation of weights of mixed indexes due to the small amount of calculation and higher consistency, therefore, the PA-FUCOM combination weight calculation method has a better performance in the differentiation of evaluation results and the consistency of evaluation results.