<p>Along with the wide application of intelligent systems in various fields, the combination of data fusion and knowledge graph has become the key to enhance the system’s problem solving capability. However, existing data fusion methods still face challenges when dealing with multi-source heterogeneous data, especially in how to effectively combine knowledge graph. Therefore, this paper systematically reviews existing data fusion methods based on knowledge graph and classifies them into three categories: fusion of raw data, fusion of raw data with knowledge graph, and fusion of knowledge graphs. Each category of methods is described and analyzed in detail by combining a general framework with specific examples. In addition, this paper also discusses the future research direction of data fusion based on knowledge graph, and analyzes the challenges and opportunities it faces. This paper provides a theoretical framework and practical guidance for the problem of multi-source heterogeneous data fusion, and provides methodological support for the development of intelligent systems.</p>