Research On Performance Evaluation Of Science And Technology Input Projects Based On SAHP-Cloud Model

The development of science and technology directly determines the level of a country’s industrial economy.Moreover,increasing investment in science and technology is the premise and guarantee for improving the level of science and technology.In addition,it is the key to ensure the long-term healthy development of the country’s economy,and it is an important manifestation of the strength of the entire country.Government financial technology investment is an important part of science and technology investment,and plays a guiding and driving role in other science and technology investment.In recent years,the government’s fiscal and technological investment has grown rapidly,and the number of science and technology investment projects has increased,providing an important and powerful guarantee for technological innovation.However,the actual effect of these inputs has become an issue of increasing concern to the public.Therefore,it is important to conduct effective performance evaluation of science and technology investment projects.This paper,from the perspective of the government,aims to improve the scientific,rationality of the performance evaluation of science and technology input projects,and conducts research on the relevant theories and existing methods of performance evaluation of science,and technology input projects.It aims at the subjectivity of current expert selection.Strong,index system construction and performance evaluation methods do not consider the randomness of expert evaluation,and propose the application of content-based recommendation algorithm,stochastic analytic hierarchy process,principal component analysis,cloud model and other theoretical optimization evaluation mechanism.The main research contents of this paper are as follows:(1)The process of constructing the indicator system is discussed by taking the application-oriented technology input project as an example.Firstly,the hierarchical network index system model is used to complete the initial construction of the indicator library following the principle of index construction.Secondly,based on the principal component analysis theory,the information sensitivity model is used to achieve the index screening,which reduces the redundancy of the indicators and the objective weight of the indicators.In view of the fact that the traditional AHP does not take into account the randomness of the expert scoring process,it proposes to use the stochastic analytic hierarchy to calculate the subjective weight.Finally,the multiplicative synthesis method combines the subjective weight with the objective weight to obtain the comprehensive weight,which maximizes the scientific and reasonable index system.(2)The method and process of performance evaluation of science and technology input projects were designed in detail.Firstly,the accuracy of the expert history review is integrated,and the content-based recommendation algorithm is improved to achieve the most suitable expert for the project to be evaluated.Secondly,experts are invited to score the indexed items according to the indicators.The fuzzy model of the cloud model is used to convert the expert’s fuzzy evaluation into quantitative data represented by three parameters:Ex,En and He,and identify and correct the evaluation of "abnormal"data.The expert weights are then determined using the "uncertainty" and "deviation" of the expert review data.Finally,using the virtual cloud generation algorithm,after assembling the expert evaluation information,the bottom layer index evaluation information is assembled into the final evaluation result.(3)Verify the scientific and rationality of the method described above by taking an application case into the actual case of project performance evaluation.