| With the rapid development of science and technology,policies,regulations and management methods related to scientific research project management have been formulated.In the management of scientific research projects,the performance evaluation of scientific research projects is very important.Scientific research projects have the characteristics of diversity and complexity,so its evaluation can not only rely on quantitative indicators.In order to fully tap the potential value and benefit of the project,expert evaluation can not be ignored.However,there is a risk of lack of fairness and impartiality in expert evaluation to some extent.Secondly,how to gather experts' opinions is also a work worthy of discussion.Based on the above problems,this paper takes scientific research projects and evaluation experts as the research object,applies text mining to the matching of scientific research projects and evaluation experts,and realizes the automatic matching of scientific research projects and evaluation experts.Firstly,the matching principle is proposed according to the selection principle of experts.On this basis,the text clustering technology and matching model are integrated to obtain the expert cluster,which effectively reduces the calculation amount of the algorithm.Then we use the text similarity calculation to achieve the matching of expert candidate set.Through the comprehensive score including the academic influence of experts,social relations and the degree of agreement with the project field,we comprehensively and comprehensively screen the candidate experts.In order to meet the needs of the model,we use segmentation,part of speech tagging,keyword extraction and other technologies to obtain expert features and project features as the input of the model.At the same time,combining with the fuzzy AHP,this paper puts forward a double hesitation analytic hierarchy process,which makes the gathering of experts' opinions more efficient and accurate.Experts can use multiple membership and non-membership degrees to fully express their hesitation information when evaluating projects,so as to make the evaluation more flexible and efficient.Then,using the fuzzy theory,a method of automatically repairing the inconsistent expert score is put forward,so as to get the acceptable consistent result,so that the inconsistent result does not need to be returned to the expert for the second review,so as to improve the efficiency of the review,and on this basis,the flow of using this method to determine the optimal scheme is proposed.Finally,an example of a specific scientific research project is used to automatically assign evaluation experts and determine the weights of evaluation indexes by using the analytic hierarchy process of double hesitation.The application scenarios of expert matching based on text mining and determining the weights of indexes by using fuzzy AHP are illustrated.The feasibility and effectiveness of the method are verified by comparative analysis.The research project and expert matching model based on text clustering proposed in this paper is not only of practical significance to avoid the problem of "relationship and circle" of evaluation experts,but also to enhance the public fairness of expert selection.Because of the similarity calculation based on text clustering,it also has the theoretical significance of reducing the amount of computation and improving the efficiency of matching.When the subjective evaluation of experts is transformed into objective value,the fuzzy theory of double hesitation can be used to deal with the indecisive judgment of experts,keep the integrity of information to the greatest extent,and make the evaluation more flexible and more extensive.The proposed method can automatically repair the expert's inconsistent score,effectively avoid the disadvantages of traditional manual correction,reduce the workload of experts,improve the evaluation efficiency,and provide reference for multi-objective decision-making problems. |
PDF Full Text Request