In the paper the problem of spreading fake news in social networks is considered. Various models for detecting such news were examined, which showed a loss of performance over time due to their dependence only on textual features. The main result of the work is the developed model for detecting fake news in the social networks, which is based on the use of knowledge graphs. The article describes in detail the structure and functionality of the proposed knowledge graph. To test the effectiveness of the proposed graph-based model, a comprehensive study of its properties was carried out and the results were compared with the results obtained using other models for detecting fake news. The comprehensive software package for fake news detection in the social networks was developed and implemented.