Research On Expert Matching Recommendation System Based On Collaborative Deep Learning

With the continuous advancement of China's informatization construction,in order to speed up the process of information reporting,project application,and process approval,the information management systems for science and technology projects around the country have been put into operation in succession,effectively strengthening scientific research management and realizing efficient management of fund projects.Transformation,standardization and fairness.In the process of scientific and technological project management,it is an important task to use information technology to find experts for evaluation of the science and technology projects to be evaluated.However,with the increase in the number of science and technology plan projects to be reviewed,this type of work is difficult to meet the timeliness requirements.For the needs of automation and rapid selection of review experts in the review of scientific and technological projects,this paper mainly did the following work.1.Investigate the current review process for science and technology projects,and clarify the urgency of recommending system requirements for the review of science and technology projects.This paper studies the recommended algorithms used in the current mainstream recommendation system,analyzes the shortcomings in the demonstration,and demonstrates the rationality of using the collaborative deep learning recommendation method to recommend evaluation experts for the science and technology projects to be evaluated.2.Based on the principle of peer review of scientific and technological project review,the National Natural Science Foundation of China shall use the uniqueness of the subject code and use the longest priority code dynamic matching method to match the research fields of the science and technology projects to be evaluated.Analyze the deficiencies of this method.3.On the basis of the highest priority subject code,combined with the principle of collaborative deep learning recommendation algorithm,the expert review information and the science and technology project summary information are modeled,and a collaborative deep learning expert matching recommendation system framework based on the science and technology project management system is constructed.According to the expert information in the review expert database,an objective and fair scoring mechanism is provided for the science and technology project review experts.The design experiment demonstrates the applicability and effectiveness of the collaborative deep learning expert matching recommendation system in recommending evaluation experts for the science and technology projects to be evaluated.This paper initially solved the problem of automatic recommendation matching between the science and technology projects to be reviewed and the evaluation experts.The research in this paper is a meaningful attempt to reduce the selection of science and technology project review experts for science and technology project managers.