What influence digital inclusive finance from policy tones? Based on machine learning

• Machine learning is introduced in investigating what impact the digital inclusive finance in China. • Different from previous researches, we use text analysis and construct the policy tone matrix to measure policy tones. • We find the notification and advisory environmental policies significantly boost the development of digital inclusive finance. This paper investigates the relationship between regional policy tones and the development of digital inclusive finance (DIF) in China. First, we employ the textual analysis method to collect policy text data from 335 prefecture-level cities in China, and then construct a policy tone matrix. Next, we use machine learning methods, including BP neural networks, LSTM, and decision trees, to simulate the actual relationship between policy tones and DIF, thereby clarifying the complex interplay between the two. The results show that the environmental policies of the notification and advisory tones significantly contribute to the development of DIF. Specifically, the advisory environmental policies have a substantial impact on the coverage breadth of DIF. Meanwhile, the economic policies, including notification, regulatory, and advisory tones, significantly promote the development of DIF. This study contributes to the existing literature on the factors influencing DIF and provides valuable insights for practical policy-making.