Abstract: To explore the research progress and development trend of machine learning technology application in soil erosion field study, CiteSpace and other bibliometric tools were used to analyze the research progress, hotspots, author's cooperation network, and future research direction and development trend of machine learning technology in this field, based on the relevant documents included in the Web of Science(WOS)core collection database.The results show that the research results in this field have been increasing exponentially since 2014.China has the largest number of publications and citations, but the intermediary centrality is lower than that of Iran and the United States.Erosion sensitivity analysis is a hot issue in this field.Most of researchers develop efficient erosion prediction models based on the faster and more accurate characteristics of machine learning compared with traditional models.Deep learning and various regression algorithms are the most commonly used machine learning methods.In the future, researchers should give full play to the characteristics of various types of machine learning, explore the latest prediction performance of deep learning, improve the prediction accuracy of soil erosion under complex environmental conditions, and reveal the contribution of main impact factors and the relevant mechanism between factors.