Institutional Collaboration and Competition in Artificial Intelligence

Abstract

The institutional collaboration and competition in academia have benefited the development of science, with inter-institutional scientific work promoting the exchange of ideas and competing fields developing rapidly. However, understanding of how the institutions collaborate and compete in science is sorely lacking, especially in emerging fields. Artificial intelligence is such a booming field currently, changing the way we live and work daily. To illustrate the problem, we try to reveal the evolution of institutional collaboration and competition in artificial intelligence by applying AI 2000 from the perspective of Science of Science. In this paper, we make multiple multidimensional statistical analyses by scrutinizing the collaboration network, research interests, talent flow, etc. We demonstrate the collaboration evolution in this field and find the advantage of inter-institutional collaboration is growing over time for papers that have been published more than 5 years. We discover the common cooperation modes of top institutions and visualize their closer cooperation. We highlight the critical resources competition among institutions in three dimensions and learn the recent trends in the field. In particular, we are concerned about the competition among institutions for cross-industry cooperation and notice the consistency of competitiveness and cross-industry collaboration. The research of this paper may support further research studies on institutional collaboration and competition as well as policy proposals for promoting scientific innovation, research management, and funding.

Keywords

• Artificial intelligence
• Science of Science
• cooperation and competition
• data analytics
• data science