logo
homeFaqIcon

AI Institute and Author Rankings by Publications

AIRankings (https://airankings.org/) is a metric-based ranking that provides insights into Artificial Intelligence research capability worldwide at various levels (e.g., authors, institutes, cities, countries). There are two indicators in the AIRankings to operationalize the AI research capability: Adjusted Publications and AI Index. While Adjusted Publications is based on the number of publications at top AI venues in core AI areas, AI Index adopts a more holistic and multidisciplinary view of AI research by calculating the geometric average of Adjusted Publications for selected areas.

  • Core Areas: AIRankings includes the following six core areas of AI: Computer Vision, Natural Language, Machine Learning, Cognitive Reasoning, Robotics, and Multi-Agent Systems. In addition, we also take into account two areas: Simulation for training AI agents in graphics environments like VR and AR; and AI In General that encompasses a broad spectrum of fields in AI, which is not equivalent to the concept of artificial general intelligence. These eight areas play in synergy and have been integrated towards building general AI systems.

  • Author Scores: Each author has two scores.Adjusted Publications is the author's total publications in selected areas, adjusted by two factors: an article is weighted by the importance of its venue, and an article with K senior co-authors (not counting students) will give 1/K score to each senior co-author. AI Index is the geometric average of adjusted publications for each selected area. It measures the breadth of an author, and gives a higher score to one with publications across multiple AI areas than another with the same number of publications focusing on a single area.

  • Institute Scores: An institute's Adjusted Publications and AI Index are similar to those of an author, but including the publications of all its senior authors. When an author changes affiliations, his/her historical scores move with him/herself to the new institute. This design is based on the assumption that the level of research of an institute is determined by its current talents, rather than historical achievements.
Selection CriteriainfoIcon
#
Institute
Adjusted Publications
AI Index
  • 1
    us 
    Carnegie Mellon Universitychart
    117 authors
    2045.5
    145.59
  • 2
    cn 
    Peking Universitychart
    122 authors
    1899.9
    120.60
  • 3
    cn 
    Tsinghua Universitychart
    96 authors
    1579.5
    101.35
  • 4
    us 
    Massachusetts Institute of Technologychart
    65 authors
    1464.0
    107.22
  • 5
    us 
    Stanford Universitychart
    54 authors
    1418.3
    90.58
  • 6
    us 
    Univ. of California - Berkeleychart
    61 authors
    1314.0
    61.22
  • 7
    cn 
    Zhejiang Universitychart
    79 authors
    1170.6
    48.25
  • 8
    kr 
    KAISTchart
    58 authors
    1126.5
    51.32
  • 9
    uk 
    University of Oxfordchart
    42 authors
    1050.6
    76.05
  • 10
    cn 
    Chinese Academy of Scienceschart
    50 authors
    1043.6
    40.41
  • 11
    sg 
    Nanyang Technological Universitychart
    47 authors
    1009.8
    51.66
  • 12
    us 
    University of Maryland - College Parkchart
    54 authors
    999.1
    88.27
  • 13
    us 
    Univ. of California - San Diegochart
    64 authors
    982.0
    66.69
  • 14
    ch 
    ETH Zurichchart
    27 authors
    932.3
    49.46
  • 15
    ca 
    University of Torontochart
    58 authors
    886.4
    71.90
  • 16
    cn 
    Shanghai Jiao Tong Universitychart
    48 authors
    883.2
    38.59
  • 17
    jp 
    University of Tokyochart
    49 authors
    855.2
    44.68
  • 18
    us 
    University of Washingtonchart
    52 authors
    853.0
    39.94
  • 19
    de 
    TU Munichchart
    24 authors
    826.5
    28.33
  • 20
    us 
    Univ. of Illinois at Urbana-Champaignchart
    56 authors
    804.4
    56.71
Contact Us