A Review of Research on Artificial Intelligence Life Cycle Based on Bibliometrics

Abstract

This paper uses bibliometric method and knowledge graph visualization technology to analyze the 237 papers of CNKI core journals from 2006 to 2021, from the key words, number of papers, authors, publishing institutions and high-impact journals. Statistical analysis, explaining the research status and research hotspots of artificial intelligence life cycle, and expounding the shortcomings and trends of research. The research on artificial intelligence technology is divided into exploratory stage and development stage according to the annual publication volume. From 2006 to 2015, the research hotspots mainly focus on "artificial intelligence" and " neural network ". From 2016 to 2021, the research hotspots mainly focus on three aspects: "artificial intelligence", "artificial intelligence technology" and "deep learning", and the research is gradually deepened, with a total of 223 journal articles. The research of artificial intelligence technology is in the development stage, and various fields are actively studying artificial intelligence technology, but the existing research focuses on the application level, the deep learning theory is not perfect, the basic technology and basic theory are ignored, and there is a lack of solutions to the problem of privacy leakage. Future research should pay more attention to basic technology and innovative research. There may be broader research space for research from two aspects: "integration and breakthrough of deep learning theory" and "machine learning evolution towards distributed privacy protection".