Thursday, July 10, 2008

Canada the most efficient producer of Computer Science research papers?

A recent article[1] looks at scientific research competitiveness of world universities in computer science, examining at individual university computer science departments as well countries as aggregates of their universities.

The study looks at 233 university computer science departments, 127418 published papers, 468244 citations, 1856 highly cited papers and 57 hot papers over the last 10 years using data from Thomson.

The evaluation criteria are (see the paper for more details):

Canada as an aggregate region came in #4 after the USA, UK, Germany, followed by Italy. Canada had the third highest paper production, citations, and "hot" papers!

For individual universities, only one Canadian university is in the top 23, with University of Toronto coming in at 16th. You can see the top 23 universities listed below:

  1. Stanford
  2. MIT
  3. UC Berkeley
  4. U Texas
  5. U Illinois
  6. Carnegie Mellon
  7. UCSD
  8. Georgia Inst Technol
  9. U Maryland
  10. Eth Zurich
  11. Technion Israel Inst Technol
  12. Princeton

  1. U Washington
  2. Tel Aviv University
  3. Harvard
  4. U Toronto
  5. UCLA
  6. UMichigan
  7. U So Calif
  8. U Minnesota
  9. Natl Univ Singapore
  10. Columbia
  11. Cambridge
I decided to do a little analysis with the numbers from this report and to look at the efficiency of production of computer science research, as measured by papers produces per capita for the above top 5 countries (as reported in this paper). Using the population figures from the CIA World Factbook and the CS paper production rates reported, I get the following rankings:
  1. Canada: 23 x10-3 CS papers per capita (7520 papers / 33,390,141 pop.)

  2. USA: 17 x10-3 CS papers per capita (51221 papers / 301,139,947 pop.)

  3. UK: 14 x10-3 CS papers per capita (8569 papers / 60,776,238 pop.)

  4. Italy: 10 x10 -3 CS papers per capita (6160 papers / 58,147,733 pop.)

  5. Germany: 8 x10-3 CS papers per capita (6779 papers / 82,400,996 pop.)

It would seem that Canada is the most efficient country with respect to per capita production of Computer Science research papers among the top five Computer Science research countries (as measured by the criteria of, and data from, this paper). So Canadians - who are continuously harangued about their great inefficiencies in all sorts of things science and technology - should at least feel somewhat good about these results. :-)

While this data is from a peer-reviewed journal article, these results and these metrics should be viewed cautiously, perhaps even with a little salt... ;-)

[1]Ma, R., Ni, C., Qiu, J. (2008). Scientific research competitiveness of world universities in computer science. Scientometrics, 76(2), 245-260. DOI: 10.1007/s11192-007-1913-7


Andre Vellino said...

Interesting! A few months ago I looked at Scimago a portal that includes data about the journals and country scientific indicators developed from the information contained in the Scopus database. I noted that "In the rankings by country over the past 10 years, Canada article citation ranking is consistently 7th by absolute numbers. On a per capita basis Canada is 6th in cited publications, ahead of the U.S., Germany and France; #1 and #2 per capita are Switzerland and Sweden."

Daniel Lemire had a closer look at the statistics and noted that the trend for Canada's overall science publications (per capita) are getting better.

Glen Newton said...


Daniel's results and these results are not necessarily that different: he looks at (what I believe) is a superset of this data: overall science publications in Canada vs. CS publications.

The per capita numbers are consistent: remember I was looking at "What are the per capita article production rates of the top 5 most competitive countries". As per this study, Switzerland and Sweden are not in this top 5, although it is likely their CS rates are similar to their general science rates that Daniel looks at.


Paul said...

Interesting that you didn't include Israel in your per-capita figure, I would guess they'd give us (I'm at UBC) a run for our per-capita money. Two universities in the top and a population of only 7.3million. I can't seem to access the article right now, but I'd be interested in that figure.