Expert secondary recommendation method and device based on improved PageRank algorithm

Abstract

The embodiment of the invention discloses an expert secondary recommendation method and device based on an improved PageRank algorithm, and the method comprises the steps: determining a scientific research association relation between each to-be-recommended expert and an author associated with the to-be-recommended expert according to an author association information set, and enabling the to-be-recommended expert to be a primary expert recommendation result; introducing the scientific research incidence relation into an original PageRank formula to obtain a new PageRank formula, obtaining a PR value of each expert to be recommended based on an improved PageRank algorithm, and taking the PR value of each expert to be recommended as a scientific research incidence relation value of each expert to be recommended; determining a weighted scientific research association relationship value of each expert to be recommended according to the affiliation relationship between the applicant and the author set; and taking the weighted scientific research association relationship value of each to-be-recommended expert as an association relationship value of each to-be-recommended expert, and determining a recommended expert according to the association relationship value. According to the embodiment of the invention, fairness and justice of expert recommendation results are improved.

Description

Technical field [0001] The embodiments of the present invention relate to the field of data processing, and in particular to an expert secondary recommendation method, device, equipment, and storage medium based on an improved PageRank algorithm. Background technique [0002] With the advent of big data, cloud computing, artificial intelligence and the Internet of Things, science and technology have developed rapidly. In this case, the number of scientific research fund project applications and paper submissions has also increased. To this end, a large number of experts need to be selected to review the declared scientific research fund projects and paper submissions. The fairness of the evaluation results of scientific research fund projects and paper submissions depends largely on whether the recommended experts are reasonable. [0003] At present, semi-manual methods are mainly used in reviewing expert recommendation. Specifically: using reviewing expert recommendation algorit...

AI Technical Summary

Problems solved by technology

[0004] However, there are at least the following problems in the prior art: On the one hand, since the review expert recommendation algorithm determines the experts to be recommended based on the similarity between the matching expert information and the project name, keywords, and field of the application, there is no Consider or seldom consider the distance between the applicant and the experts to be recommended. Generally, review experts tend to favor applicants who are close to each other, so the review results from the experts to be recommended determined by the above method may be biased to a certain extent. , therefore, will reduce the fairness and rationality of the project review results; on the other hand, due to the subjective limitations of user judgments and possible management loopholes, the unfairness of the screening results may further increase

The above shows that using the existing method to determine the recommended experts from the experts to be recommended will reduce the fairness and impartiality of the expert recommendation results, which will lead to the accuracy and rationality of the review results of project and paper submissions.

Images