How does enterprise digital transformation impact green innovation performance? A machine learning-based study

Purpose The purpose of this paper is to explore the differences in the impact mechanism of different types of enterprises’ digital transformation on green innovation performance. Design/methodology/approach From the perspective of production and operation, this study deconstructs enterprise digital transformation into six characteristics and measures these characteristics using data from listed industrial companies between 2016 and 2020. Various machine learning methods, such as cluster analysis and Bayesian networks, are applied to explore the complex impact mechanism of digital transformation characteristics on green innovation performance. Findings The findings are as follows: (1) Enterprises with different digital transformation modes exhibit distinct characteristic differences. (2) Different digital transformation modes have diverse impact mechanisms on green innovation performance, and higher levels of digital transformation correspond to more complex impact relationships. (3) Emphasizing digital transformation innovation activities within enterprises has a positive ripple effect on green innovation performance. (4) Digital asset input has a limited influence on green innovation performance and exhibits varied effects on enterprises at different levels of digital transformation. Originality/value The conclusions of this research will help enterprises understand their digital competitiveness and guide their digital transformation practices to enhance green innovation performance. These findings will also assist governments in formulating policies that promote the development of green innovation in the digital economy era.