Evaluation Method Of The Node Importance Within The Scientific Collaboration Network

In the research works, the researchers will work together to set up a research team in all kinds of ways, scientific collaboration has played as a key role on the development of science. The key researchers in the teams which are constructed by collaborations play as the academic leaders and promote the stability, sustainable development of the teams. Through the academic cooperation and communication, they will improve the innovation capabilities within the teams. To identify and evaluate this personnel is the key to the assessment and management of the knowledge workers in the organization, and the evaluate results can be used to implement the brain drain crisis management.A scientific collaboration network is constructed by using the collaborated papers'data information, in which the nodes are presented by researchers and the edges are presented by collaborations between them, and the study of evaluating the key researchers can be constructed to be a model of important nodes'measurement in the scientific collaboration network. Based on the practical significance of the scientific collaboration, the evaluation method of the node importance is proposed in two different aspects, which are weighted on nodes and edges and influence of nodes, respectively. Through the application of the method on the empirical data, some valuable results are gotten. In this paper, several research works have been studied as following:Firstly, the weights'influence on the node importance in the weighted networks is deeply analyzed. Based on this analysis results, a method of calculating the nodes'and edges' weights is put forward. The nodes'weights are determined by the collaboration contributes, and the edges'weights are determined by the amount and efficiency of collaborations which are represented by the papers'number and academic values. By synthesizing those factors, the method of evaluating node importance is put forward.Secondly, from the aspect of influence ability, the factors which impact the node importance are further analyzed. Under the different relationships between nodes and their neighbors, the direct influence and indirect influence of the node are measured. Based on the amplification effect of the influence, the impact factor of the node is measured. Combining with the weights information, an comprehensive evaluation method of node importance is proposed.At last, a practical scientific collaboration network is constructed by using collaborated papers of a research institution in management science field, and then, the method which has been proposed is tested. In the process of the testing, the topology of the practical network is analyzed by the visualization results which are intuitive description of the impact of the important nodes on the network. After that, the rank of the important nodes is given out through the practical numerical calculation, and it is contrastively analyzed with the real condition of the scientific collaboration in this institution. On this basis, the measures of improving the researchers' research capabilities are given out as the aspect of the management of key personnel in the research institutions.