Data-Driven Research On Group Collaboration Efficiency Model And Application

In recent years,with the development of Internet-related technologies and the arrival of the era of big data,the amount of data information has increased exponentially,providing a good data foundation for large-scale group collaboration related research.The data-driven group collaboration efficiency model studied in this article selects the fund research team that has successfully applied for the management science and engineering disciplines under the management science department since the establishment of the National Natural Science Foundation of China as the research object.Based on the relevant scientific research data crawled on the Web of Science research.The main work and achievements are:(1)Design of group collaboration indicator systemThe construction of group collaboration index system is the basis for evaluating the group collaboration efficiency of scientific research teams.The research in this article is data-driven and comprehensively considers all aspects that affect the collaboration efficiency of scientific research team groups.It divides the data from three first-level dimensions: basic information dimension,work capability dimension,and interpersonal communication dimension.A set of qualitative and quantitative combinations is designed.It comprehensively describes the group collaboration characteristic attribute index system,and analyzes the relevant index data of the management science and engineering research fund project team,which provides data and theoretical support for the subsequent construction of the scientific research team group collaboration efficiency model.(2)Construction of scientific research team group collaboration efficiency modelThis paper studies the management science and engineering research project team that has successfully applied for the National Natural Science Foundation of China,and evaluates the collaboration efficiency of the scientific research team based on the constructed scientific research team group collaboration index system,so as to obtain the four high-efficiency characteristics unique to the high-efficient scientific research team collaboration Attributes.Based on the index data of the scientific research team of the Management Science and Engineering Fund project,first,the high-efficiency index is determined according to the index that can accurately divide the collaboration efficiency of the scientific research group from the many indexes;then,the statistical analysis method is used to analyze the index.Finally,by analyzing the common characteristics of high-efficient scientific research teams,a model of scientific research team group collaboration efficiency is constructed,which describes the four high-efficiency characteristic capabilities of high-efficient scientific research teams,namely,the impact of results,the ability of continuous innovation,and the ability to stabilize output and the ability to solve problems.(3)Verification of group collaboration efficiency modelIn order to verify the validity and scientificity of the constructed group collaboration efficiency model,six machine learning classification algorithms were used to construct a group collaboration efficiency prediction model.It can be seen from the results that the group collaboration efficiency model proposed by the research team can effectively evaluate the cooperation effect between members in the group.By combining the evaluation results of various indicators,it is concluded that the Adaboost prediction model performs well in predicting the collaboration efficiency of scientific research team groups.At the same time,the high-efficiency index identified in this paper can effectively evaluate the efficiency of scientific research team collaboration through the prediction model of scientific research team group collaboration efficiency,which proves that the data-driven group collaboration efficiency model proposed in this paper is feasible.