Abstract: From the perspective of National Natural Science Foundation projects, this study delves into the theme and direction synthesis of research in the field of Information Resources Management, aiming to grasp the trends in this discipline as it undergoes transformation. A framework for analysing the developmental trends of the discipline has been constructed, integrating theme mining and deep learning models. Based on the NSFC's G0414 project data from the past five years, this study conducts visual analyses of project theme identification, differences between application and funding themes, thematic association structures, and their evolution over time.The funding ratio for the Information Resource Management discipline remains stable and at a high level, primarily concentrated among a few top-ranking universities. A limited number of thematic areas have received final funding, including information demand, multimodal computing, knowledge discovery, intelligent empowerment, large models, and risk management. These themes exhibit varying degrees of interconnection and influence, characterized by ongoing evolutionary traits. In the realm of natural science research, the overall development of the Information Resource Management discipline is promising, with stable funding levels focusing on specific research directions. However, there is a significant discrepancy between project applications and funding outcomes.

Key words: Information resources management, Developmental trends of the discipline, Natural science foundation projects, Theme mining, Content representation models