In order to intuitively understand the development status and research frontier of artificial intelligence, analyze the similarities and differences between domestic and foreign research, and help domestic artificial intelligence research. Based on the journal papers from 2008 to 2019 in Web of science database and CNKI database, scientific knowledge mapping and visual analysis of journal papers are carried out with Citespace software. According to the objective data and the map of scientific knowledge, it is found that after 2016, the field of artificial intelligence ushers in a new upsurge, and presents a pattern of “China and the United States”. In terms of the quality of published papers, North America is currently the region with the highest level of artificial intelligence research. At present, the main force of artificial intelligence research is colleges and universities, and the system of combining production, teaching and research has not yet formed. The research topics have distinct characteristics of the times, and artificial neural networks, algorithms, big data, robots, computer vision, legal ethics and so on have become the current research hotspots. Finally, according to the evolution of artificial intelligence research context and high-frequency words, three research frontiers of this field, namely “deep reinforcement learning”, “artificial intelligence +” and “intelligent social science”, are put forward to provide direction suggestions for the follow-up artificial intelligence research.