Abstract: Since entering the digital economy era, the negative effects caused by the "algorithmic black box" have become increasingly prominent in China. For this reason, China has put forward new requirements for digital platform companies to improve the transparency and interpretability of algorithms. Based on this, a digital platform companies optimizing algorithm transparency strategies diffusion model has been developed by using complex network evolutionary game and the following conclusions were drew in this paper. Firstly, government regulation plays a key role in whether digital platform companies are choosing to optimize their algorithmic transparency strategies. However, with the number of digital platform companies increasing, the government regulation and penalties will be less effective. Secondly, the supervision of consumer plays a supporting role and will be particularly more important as the number of digital platform companies becomes larger. Thirdly, the cost of optimizing algorithms and the risk of disclosing the algorithms are both important factors that prevent digital platform companies from optimizing their algorithm transparency, and the ability of companies to resist the risk will also decrease as the number of digital platform companies grows. Finally, counter-suggestions were put forward based on the conclusions, which may provide some reference to the governance of digital platform companies’ “algorithm black box” in China.

Key words: Digital platform enterprise, Governance of algorithm, Algorithmic black box, Algorithm transparency, Complex network evolutionary game