With the development of mobile applications and location technology, people can share their location and conduct various social activities by mobile phones, which promotes the formation of location-based social network(LBSN). Data from different sources with different formats are generated in LBSN. How to extract knowledge from these data and fuse them together is of great significance to analyze the characteristics of user in LBSN. In this paper, an ontology-based LBSN domain knowledge graph construction method is proposed. Firstly, the data model and ontology of the graph is defined, which includes 8 entities, 11 relations and their related attributes. Then, the construction process of the knowledge graph and the process to merge knowledge extracted from check-in data and the label data is defined. Finally, a case study is conducted to verify the effectiveness of the method. Result shows that the method can construct knowledge graph for LBSN and merge knowledge extracted from multi-source effectively, and the knowledge graph constructed can promote the storage of LBSN domain knowledge and user feature analysis.