Design And Implementation Of Collaborators Recommendation System Based On Project Collaboration

Modern disciplines have emerged a new trend of integration and cross-cutting.It is becoming more and more difficult for single people to complete scientific research activities,and scientific research cooperation is very important.Research collaboration can facilitate communication between researchers,consolidation and efficient use of research resources,and help to improve the efficiency of scientific research,scientific innovation and research quality.For researchers,how fast,high-quality scientific research to find a suitable partner becomes a problem to be solved at present.On the other hand,with the continuous development of scientific research management information in our country,we have accumulated a wealth of scientific research datas,how to explore and use existing research datas to solve this problem is a worthy of discussion.In this paper,based on the authoritative and rich scientific research datas,this paper proposes a set of cooperative recommendation model based on project collaboration relationship,and constructs a simple,efficient and extensible collaborators recommendation system.In this paper,the project cooperation relationship among researchers in scientific research management data resources is abstracted as weighted and undirected network,and the weight calculation model is put forward considering the multi-dimensional influencing factors such as the number of collaboration,the order of collaboration signature and the cooperation year,greatly improving the weight of the dimension of the problem.In order to solve the problem that the traditional node similarity model is not enough and the complexity is too large,a similarity model based on locally variable-length path cooperative relation is proposed.The concepts of coincidence degree and reconsideration degree in the research field of researchers are put forward in this paper,and the similarity model of the research field based on the coincidence degree and reconsideration degree of research in the research field is created.Finally,it is helpful to improve the recommended accuracy by combining the similarity model of cooperative relationship and the similarity model of research field.The system uses a service-oriented architecture,with high ease of maintenance,scalability and loosely coupled.The recommendation system provides social science researchers with the search for potential partners,which greatly solves the problem of the efficiency and quality of the collaborators seeking to improve the utilization of social science data resources and the use of social science data a new direction.Based on the theory of network science and recommendation system,we can analyze the generation and evolution of cooperative relationship network and the mechanism of group and community development.By analyzing the factors such as node degree and importance,we can apply to the core and author of the recommendation.The recommended model has high exploration value both in theoretical research and application.