China currently set to lead the world in AI research

Academics in the AI field are leaving the US and Europe’s academic sectors to work in industry, but China is on track to become the biggest source of AI research globally

China is growing in importance as a global leader in artificial intelligence (AI) research, and there is an increasing trend of researchers moving from academia to industry, especially in the United States, according to new a study released by information analytics business Elsevier.

The report shows that, globally, AI research has accelerated, growing by more than 12% annually in the past five years (2013-2017), comparing to less than 5% for the previous 5 years (2008-2012). By contrast, research output overall, globally across all subject areas, has grown by 0.8% annually over the past five years (2013-2017).

The analysis finds that industry in the United States attracts the most AI talent from both local and international academia. In Europe there’s a stronger move of academic talent moving to non-European industry.

Over the last three years, the data shows Chinese academia attracting more AI talent than it is losing, confirming that the country is on track to establish a leading position in AI research. Having overtaken the United States in AI research output in 2004, China is set to overtake Europe and become the biggest source of AI research globally in four years, if the pace of current trends continues.

The report is not a conclusion, but the start of a discussion on how we best enter the era of AI and increasingly symbiotic technology.
– Dan Olley, CTO, Elsevier

Another important revelation is that despite the increasing societal reference of AI and the media attention on the ethical implications of AI, academic research on the ethics of AI has been limited.

Dan Olley, Chief Technology Officer at Elsevier, said: “The new generation of technologies, commonly umbrellaed as AI, are so important and yet, there appears to be no shared understanding of its exact definition. With this comprehensive study of research performance in AI we aim to provide clarity on and insights into the field’s dynamics, trends and parameters. The report is not a conclusion, but the start of a discussion on how we best enter the era of AI and increasingly symbiotic technology.”

Enrico Motta, Professor of Knowledge Technologies at the Open University in the UK, expert contributor to the report, said: “This report applies extensive text mining and semantic analytics across literature from different sectors to uncover how to more comprehensively define the AI field – essentially using AI to map AI. It is the most comprehensive characterization of AI outputs across different sectors delivered so far.”

Reviewing 600 documents and over 700 field-specific key words across four sectors – research, education, technology, and media – the semantic analysis reveals that the field of AI focusses on seven distinct research areas:

  • Search and Optimization,
  • Fuzzy Systems,
  • Natural Language Processing and Knowledge Representation,
  • Computer Vision,
  • Machine Learning and Probabilistic Reasoning,
  • Planning and Decision Making, and
  • Neural Networks.

Of these areas, research in machine learning and probabilistic reasoning, neural networks, and computer vision show the largest volume of research output and growth.

Other regional findings highlighted in the report:

  • International mobility and collaboration patterns suggest that China operates in relative isolation from the wider research community.
  • Europe is the largest and most diverse (in terms of research areas within AI) region in AI scholarly output, with high and rising levels of international collaborations outside of Europe.
  • In 2017 India is the third largest country in terms of research output in AI after China and the US. Iran is ninth in publication output, on par with France and Canada. Germany and Japan remain the fifth and sixth largest in AI research output.

Data used in the report come from Elsevier’s Scopus, Fingerprint Engine, PlumX, ScienceDirect, and SciVal, RELX’s TotalPatent, and further draws on public sources, including dblp, arXiv, Stanford AI Index, kamishima.net, and Kaggle, as well as datasets provided by the Institute of Automation, Chinese Academy of Science.

Analyses were further informed by experts from around the world who advised on the report’s development, including the research questions, methodologies, and analytics, and who provided a policy context for the findings.