Abstract: 【Purpose/significance】Mining scientific knowledge in scientific literature based on knowledge element, establishing the
connection between scientific knowledge, and constructing a fine-grained knowledge graph, aiming to break through the barriers be?
tween knowledge and meet the fine-grained knowledge needs of users.【Method/process】Firstly, this paper builds a graph-based
knowledge element representation framework, and builds a fine-grained knowledge organization model with knowledge element as the
core; Secondly, design a knowledge graph for the knowledge elements of scientific literature, and explore the construction process of
knowledge graph, in order to realize automated construction of knowledge graphs; Finally, the abstracts and introductions in scientific
literature are used as experimental data to conduct empirical research, and the knowledge graphs for the knowledge elements of scien? tific literature are constructed.【Result/conclusion】The knowledge graph constructed in this paper can not only intuitively display the questions that need to be explored in academic papers, the proposed methods and models, etc., but also reveal the intrinsic relation? ship between scientific knowledge.【Innovation/limitation】This paper explores the fine-grained knowledge organization model and builds a knowledge graph for the knowledge elements of scientific literature. In future research, the knowledge graph construction pro? cess will be continuously improved, and the application fields of knowledge graph will be explored.