Abstract: [Purpose/Significance] Under the policy background of promoting the converged value-added utilization of government open data, it is crucial to elucidate the driving factors and mechanisms for establishing collaborative relationships among multi-source heterogeneous government open data. This clarification helps provide references for promoting the convergence and utilization of government open data resources across departments, levels, and regions. [Design/Methodology] The formation of the collaborative relationship network of government open data was analyzed using an explanatory framework and an Exponential Random Graph Model （ERGM）. An empirical study was then conducted based on the data source information of the innovative applications of government open data in Shandong Province. [Findings/Conclusion] The government open data collaborative relationship network presents significant clustering and transitivity structural characteristics. Government open data with high data volume, high update frequency, high download counts, high browse counts, and the same domain, administrative region, update frequency and openness level being more likely to establish collaborative relationships. Data item relationship network and provider relationship network play a cross-network effect on the formation of government open data collaborative relationship. [Originality/Value] This paper employs social network quantitative method to model and analyze the collaborative relationships of government open data. The research findings are valuable for understanding the potential mechanisms of the collaborative supply, integration, and value-added utilization of multi-source and heterogeneous government open data , as well as providing valuable insights for future research and practice.

Keywords: Government open data, Innovative use of data, Collaborative relationship network, Exponential Random Graph Model