The aim of this study was to explore the design of tumor disease diagnosis based on knowledge graph. By constructing and utilizing knowledge graph, combining tumor case data and medical knowledge, an innovative decision support system for tumor disease diagnosis is proposed to improve the accuracy and efficiency of tumor disease diagnosis. The system uses knowledge graph technology to structurally associate professional knowledge in medical field and tumor case data to form a rich knowledge network. Through the construction and updating of knowledge graph, the system can realize the comprehensive analysis and comprehensive reasoning of tumor cases, and provide doctors with more accurate and comprehensive diagnostic support. This system will also integrate data mining technology, analyze and model tumor case data, and dig out the potential rules and characteristics hidden in the data. Based on these models and algorithms, the system can predict and classify new tumor cases and assist doctors to make diagnostic decisions.