Drama communication is an important path for the inheritance and development of drama art. In order to clarify the research development trend of the innovative communication of Ji opera art in the era of Internet+, this paper combines the principles of graph theory and data mining technology, and proposes the knowledge graph relationship extraction based on graph neural network, which utilizes the BERT model to encode the input sentences, and constructs the graph structure through the Gaussian graph generator. On this basis, the knowledge graph of Jiyu Opera art innovation dissemination is constructed. The Ontology Model of the Knowledge Graph was constructed using the Protégé tool, which covered 4 event entity categories such as “Art Creation Event”, “Stage Presentation Event”, “Communication and Promotion Event”, and “Inheritance Education Event” and 9 event element entity categories such as “Time”, “Place” and “People”. In the analysis of the key words of the knowledge graph, the research hotspots and their evolution in the field of Jiqu art innovation and communication are obtained. The intensity of the keywords “digital inheritance” and “cross-border integration” reached 2.45 and 2.23, respectively, which were research hotspots in the field. The research interests of policy support and “international communication” have lasted for a long time, while “IP development” and “short video marketing” will be highlighted in 2021-2024.
