2024 CSRankings National Computer Science Rankings Released! CMU dominates the list, MIT falls out of the top 5

The 2024 CSRankings National Computer Science Major Rankings have just been released!

This year, among the best CS universities in the United States,

Carnegie Mellon University (CMU) ) ranks among the best in the United States and in the CS field, while the University of Illinois at Urbana-Champaign (UIUC) has been steadily ranked second for six consecutive years. Georgia Tech ranked third.

Then, Stanford University, University of California at San Diego, University of Michigan, and University of Washington tied for fourth place in the world.

It is worth noting that MIT’s ranking fell and fell out of the top five.

CSRankings is a global academic computer science initiative initiated by Emery Berger, a professor in the School of Computer and Information Sciences at UMass Amherst. Ranking items. The ranking is based on objective research indicators, providing a relatively transparent assessment tool for academics and students.

CSRanking’s subdivision rankings are divided into 4 major categories (27 small subdivisions), namely AI, systems, theory and interdisciplinary fields.

In the AI ​​sector, there are mainly 5 subdivisions: 1. Artificial intelligence; 2. Computer vision ; 3. Machine learning; 4. Natural language processing; 5. The Web & information retrieval.

National CS ranking: CMU firmly ranks first, MIT falls out of the top 5

3 universities CS strength is tied for the top 8 in the United States, namely MIT, University of California, Berkeley, and University of Maryland, College Park.

Then there are Cornell University, Northeastern University, Purdue University, and the University of Texas at Austin, with 5 universities tied for the top 15 and 4 universities tied for the top 20. .

Specifically, CMU is best at machine learning, NLP, vision, and AI.

Machine learning and NLP research at the University of Illinois at Urbana-Champaign (UIUC) is its greatest strength.

Georgia Institute of Technology has outstanding research in the field of ML and security.

Let’s take a look at the research performance of the four universities tied for fourth place: Stanford University, University of California at San Diego, University of Michigan, and University of Washington.

Visualization of the proportion of research fields of the three universities tied for 8th.

Ranking basis: number of papers published at top conferences

Currently, the "U.S. News" started in 1983 Ranked the most prestigious by US News and World Report.

However, the US News rankings are entirely based on reputation, relying on surveys of department heads and graduate school deans.

Take the 2023 US News world rankings, which went completely wrong. The rankings of 213 institutions were wrong, and the study abroad circle was also in turmoil.

Including the previous falsification of Columbia University’s ranking data in US News, it also caused a bloody storm.

In addition, indicators based on the number of citations are also suspected of being filled with water. For example, some universities encourage faculty and staff to cite each other, thus creating "citation cartels".

Not only that, not all paper citations are free, and they change quickly, and citation statistics systems like Google Scholar do not do a very good job with author ambiguity.

In order to provide everyone with a meaningful and transparent ranking system, Emery Berger organized the global strength ranking of colleges and universities in the field of computer science, based entirely on "Research Indicators" for ranking.

Specifically, CSRankings is measured by the number of papers published by the vast majority of college faculty at major top conferences in the field of computer science.

Top natural language processing conferences include ACL, EMNLP, and NAACL; top conferences in the field of computer vision include CVPR, ECCV, and ICCV; papers from machine learning and data mining conferences come from ICML and KDD , NIPS; top artificial intelligence conferences include AAAI and IJCAI.

It can be seen that this method is to encourage faculty and staff to publish papers at top conferences, and all codes and data of CSrankings are shared, while also preventing fraud.

What are the selection criteria for faculty and staff?

The inclusion criteria for the database are that as long as they are full-time, tenured faculty members on a specific campus and can independently provide guidance to doctoral students majoring in computer science, they can be included in the database. .

Therefore, this approach extends the coverage of the database to some faculty in other departments who have part-time contracts with the Computer Science Department or similar departments and can provide Ph.D.s in CS Students provide guidance.

Please note that full time means working at least 75% of the time during the entire academic year.

How to calculate author signature and number of papers?

A faculty member may receive 1/N points for a paper, where N is the number of authors, their affiliation or status (staff, student, or other status) Nothing to do. This number will never change.

In the case where all authors are/eventually become faculty members in the database, then a paper can only count for a maximum of 1.0 points.

If the academic output of a paper is not divided by the number of authors, but simply counts the number of papers, then the author can easily artificially manipulate and exaggerate the output of a single paper by adding more authors. Output.

To avoid this situation, the academic output of the paper must be divided. This can motivate authors to treat attribution of academic output appropriately and not abuse multiple authors to exaggerate the output of a single paper.

The above is the detailed content of 2024 CSRankings National Computer Science Rankings Released! CMU dominates the list, MIT falls out of the top 5. For more information, please follow other related articles on the PHP Chinese website!