Abstract: 【Purpose/significance】In the era of digital intelligence, the integrated research on big data security and governance has be⁃ come a current research hotspot and a key issue. This paper explores the core issues, development trends and evolutionary trajectories in this field by mining the research topics in big data security and governance.【Method/process】It combs and verifies the existing lit⁃ erature in this field. Utilizing the data from China National Knowledge Infrastructure (CNKI), for the core journal literature related to big data security and governance during the period from 2012 to 2024, an improved BERTopic topic model and a collaborative analy⁃ sis framework with GPT-4 are adopted to mine the implicit research topics and their importance in the texts, and to display the dy⁃ namic evolutionary process and trends of the topic contents.【Result/conclusion】The research finds that the field of big data security and governance presents three main evolutionary characteristics, namely the interactive evolution of theory and practice, the in-depth expansion of multi-dimensional security, and the dynamic balance between security and development. Based on these, it puts forward the prospects for improving the characteristic development in the field of big data.【Innovation/limitation】This study introduces large language models to deeply explore research topics and their evolutionary characteristics in the field of big data security and gover⁃ nance. In the future, it will focus on the field of emerging technology governance and security, aiming to achieve "virtuous governance" of big data