Abstract: Data sharing is a critical way to fully release the value of data resources. This paper starts from opening up data circulation channels between the government and platform manufacturers and promoting data sharing, constructs an evolutionary game model between platform firms and the government, studies the strategic choices of the two in the game process, simulates the equilibrium point of the evolutionary stable strategy between them in different situations, and explores the influence of data transformation ability, supervision cost, sharing cost and other factors on the sharing behaviors of platform firms. The study finds that when the social reputation coefficients of platform firms and the government increases, the distribution coefficient of government economic benefits declines, and the amount of data shared by platform firms rises, the system can evolve from “worst”, “ineffective” and “bad” states to the “ideal” state respectively, and finally achieve the Pareto optimality of both platform firms and the government; where there exists differences in the data conversion capability of platform firms and the government, only when the government has a significantly stronger data conversion capability than platform firms, will the choice of data sharing behaviors be significantly promoted; at the same time, the improvement of data conversion capability and the increase of supervision cost will also prompt platform firms to choose sharing; however the rise of sharing cost will delay or inhibit the data sharing behaviors of platform firms. Finally, this paper puts forward corresponding policy recommendations.

Key words: data sharing, friendly cooperation, evolutionary game, path analysis