Research on Construct and Relation Reasoning Method Based on Knowledge Graph

The current knowledge graphs suffer from widespread issues of structural disorder and missing data. This paper aims to address these problems by conducting research and improvements using relationship inference methods. It demonstrates the effectiveness of relationship inference methods in complementing and rectifying knowledge graphs through a specific case study. Building upon a brief introduction to the logic and methodologies involved in constructing knowledge graphs, this paper extensively investigates relationship inference methods based on knowledge graphs. The research successfully resolves the problems of structural disorder and missing data in knowledge graphs, thereby enhancing the quality and usability of the knowledge graphs. This achievement holds significant implications for further analysis and research utilizing big data, facilitating a better understanding and application of abundant knowledge resources.