Technology Topic Mining and Patent Data Analysis of NdFeB Permanent Magnet Materials Based on Machine Learning

In this study, the NdFeB patent data was analyzed using patent data analysis and BERTopic topic modeling methods to explore the application status, patent types, technology life cycle, distribution of technological topic, and technological evolution trend. The findings reveal that NdFeB patent technology is still in a phase of rapid development, with China and Japan leading in the number of patent applications and institutions globally. Among these, invention patents are predominant. The IPC technology fields of these patents focus primarily on magnet powder and its preparation, motors, surface coatings, and related areas, with enterprises driving most of the technological innovation. The BERTopic topic model identified 41 technical topic within the patent data, which were further clustered into 9 technical directions. Dynamic topic modeling (DTM) analysis highlighted that new technologies such as surface protection, waste recycling, permanent magnetic component, and grain boundary diffusion have garnered significant attention in recent years. In contrast, technologies related to grain boundary composition and raw material formulation have maintained long-term interest.