基于美国国家基金数据的人工智能前沿分析  

Frontier analysis of artificial intelligence based on national fund data of the United States

作　　者：刘城 廖林望 丁倩[1] LIU Cheng;LIAO Linwang;DING Qian(Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南京理工大学,江苏南京210094

出　　处：《中国高新科技》2021年第4期68-73,88,共7页

摘　　要：随着新一轮人工智能研究热潮的兴起,以美国为代表的世界主要发达国家率先将发展人工智能作为国家重大战略。因此从国家基金角度识别并分析美国人工智能的研究前沿,可以发掘其侧重方向、领先优势和协作关系,为我国科技政策制定提供数据支撑与决策参考。以美国国家科学基金会官网数据为数据源,按主题分类采集人工智能研究的基金数据,经清洗与区分后,利用项目负责人关键词耦合、关键主题时间演化和机构主题共现等方法,进行研究前沿的识别与分析。通过对项目负责人与机构的检测和分析,得到了美国在人工智能方向的研究前沿及侧重方向;将项目负责人分成8类,识别出每类的主要研究方向;对高产机构所研究的主题进行识别,通过综合分析识别出美国在人工智能领域的研究前沿。With the rise of a new wave of artificial intelligence research,major developed countries in the world represented by the United States have taken the lead in the development of artificial intelligence as a major national strategy.Therefore,identify and analyz the research frontiers of artificial intelligence in the United States can explore its focus,leading advantages and collaborative relationship,and provide data support and decision-making reference for Chinese science and technology policy-making.Based on the data from the official website of the National Science Foundation of the United States as the data source,the fund data of artificial intelligence research were collected according to the subject classification.After cleaning and distinguishing,the research frontier of the United States was identified and analyzed by using the methods of keyword coupling,time evolution of key topics and co-occurrence of institutional themes.Through the detection and analysis of researchers and institutions,the research frontiers and emphases of artificial intelligence in the United States were found.The main researchers were divided into eight categories,and the main research directions of each category were identified.The research topics of high-yield institutions were identified,and the research frontier of artificial intelligence in the United States was identified through comprehensive analysis.

关 键 词：人工智能 国家基金 美国国家科学基金会 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]