THE DataPoints is designed with the forward-looking and growth-minded institution in view
China has been producing almost twice as many papers on artificial intelligence as the next highest-placed country in terms of publication volume for the field, a data analysis for Times Higher Education has shown.
Data from Elsevier’s Scopus database provided to THE illustrate China’s huge drive on research in the area, with researchers in the nation notching up just over 41,000 publications from 2011 to 2015.
In terms of publication volume, the US was second during the period with almost 25,500 publications while Japan was in third place (about 11,700) and the UK in fourth (about 10,100).
However, although China scored high in terms of volume, it was only 34th in terms of field-weighted citation impact (which allows for differences in citations according to subject and year), suggesting that most of the papers were not of the same quality as those coming from the US (fourth for citation impact), for instance.
|Country||Publications||Field-weighted citation impact|
Leading the world on this measure was Switzerland, with a citation impact score of 2.71, followed by Singapore (2.24) and Hong Kong (2.00), although all three of these produced fewer than 2,500 publications on AI over the time frame.
Looking at individual institutions that published more than 500 times on AI shows that only one in China – the Institute of Automation, Chinese Academy of Sciences – had a citation impact above the world average of 1.
The list ranked by citation impact is topped by the Massachusetts Institute of Technology with a score of 3.57. This is way ahead of the rest of the chasing pack, including Carnegie Mellon University and Nanyang Technological University, Singapore.
|Institution||Country||Publications||Field-Weighted Citation Impact|
|Massachusetts Institute of Technology||United States||1,011||3.57|
|Carnegie Mellon University||United States||1,311||2.53|
|Nanyang Technological University||Singapore||1,197||2.51|
|University of Granada||Spain||587||2.46|
|University of Southern California||United States||627||2.35|
|Technical University of Munich||Germany||656||2.27|
|Institute of Automation, Chinese Academy of Sciences||China||588||2.26|
|Hong Kong Polytechnic University||Hong Kong||602||2.20|
|National University of Singapore||Singapore||807||2.14|
|Chinese University of Hong Kong||Hong Kong||530||2.09|
In recent years, it has been industry – in emerging tech areas such as self-driving cars – that has been the driving force behind the explosion in AI research, both in the West and in China, according to Alexander Wong, Canada research chair and professor in systems design engineering at the University of Waterloo.
Waterloo, the first university in Canada to receive government approval to test a self-driving car on public roads in the country, comes 11th in terms of citation impact for AI research.
“Before the recent revolutions in AI, industry had been very slow in adopting and embracing [the field], but now they…are rapidly adapting their businesses to push the frontiers of AI, and more importantly working closer to researchers and universities to support them in pushing advances in AI at an incredible pace,” Professor Wong said.
Sean Holden, a senior lecturer in computer science at the University of Cambridge, said a key factor in the current growth of AI research was that recent advances “have been in areas that can obviously be monetised” in a manner different from “traditional” uses of AI.
“The assorted boxes that you can talk to in your house – that try to assist you – have only more recently been feasible and represent a completely different way of making revenue. Such avenues are opened when notable advances are made, of which there have recently been several,” he said.
However, he warned that there was a “cycle of hype” surrounding AI which was currently “at its peak”, leading to companies and governments showing more interest. In the long term it was universities that would still have impact in driving the field forward, he said, because they were “more interested in science than generating return for shareholders”.