Abstract:[Purpose/Meaning]With the increasing complexity of research issues, academic research cooperation has become the main way to carry out research and scientific innovation activities. Through the analysis of high-performance scientific research cooperation network, we can understand the significance of scientific research cooperation in a deeper level, and guide the direction and mode of scientific research cooperation network management. [Method/Process]The objects of this study are the scientific research cooperation network built by the top five scholars in the 2017 comprehensive index of CNKI. Data samples are the domestic journal papers participated by these scholars from 2006 to 2017. Dynamic Network Analysis as an effective analysis tool is employed to construct the weighted co-authored network and Keywordnetwork, both of which edge weights are subject to the impact factors of journal papers. The potential common characteristics are revealed from three different network levels, including network overall structure, overall attribute and individual attribute. [Result/Conclusion]The results show that the high performance research cooperation network has the characteristics of large scale, high average clustering coefficient, small average path length, and large number of researchers and Keywordsflow. These research results can reflect the characteristics of the actual scientific research cooperation network, and have important theoretical significance for revealing the common characteristics of the high-performance scientific research cooperation network. It can provide scientific basis for personnel management departments and scientific researchers to introduce the strategic planning of collaborative innovation, organizational guidance mechanism, management system innovation and other measures.