A Knowledge Graph Construction Method for Legal Documents

“China Judgments Online” is a platform that collects and stores court judgments from various levels of courts nationwide, providing a vast amount of legal cases and rulings. Leveraging these information to promote legal transparency, fairness, and efficiency holds significant research value. And knowledge graph is a tool for organizing and representing knowledge graphically. By constructing relevant legal knowledge graphs, the entities, relationships, and attributes in legal data are structured to help legal practitioners better understand and analyze the relationships and characteristics between court judgments. Therefore, we propose a bottom-up approach to construct a legal knowledge graph. Firstly, we crawl relevant legal data from a wide variety of websites. Then, we employ different methods to extract entities and their relationships from both structured and unstructured data. Finally, we integrate these two sets of data through entity matching and attribute matching, and visualize the constructed knowledge graph using Neo4j.