Map of Russian Science

» see mapofscience.ru as it lived, on the wayback machine

~ the obituary ~

The tender went out at the end of 2012 with a ceiling of a hundred million roubles, and seven bidders came in. Moscow State University offered to do it for fifty. PricewaterhouseCoopers Russia bid ninety, the highest number on the table, and won, with ninety days allowed. About forty million of that went straight back out to Thomson Reuters for Web of Science data, and twenty seven went to a subcontractor called AT Consulting, who actually wrote the website. The Map of Russian Science entered trial operation on 12 November 2013.

The premise was that a country could keep an automatic register of its own scientists. Pull publications from the Russian citation index and from Web of Science, patents from Rospatent, monographs from the Book Chamber, add Scopus under a later contract, and the state would have activity indicators for every researcher and every institution without asking anybody to fill in a form. By the end of that first year it claimed around five thousand organizations and over half a million people.

The half million was already the problem. Analysts at Troitsky Variant put the Map's own claim closer to six hundred thousand and pointed out that it counted students and postgraduates, plus foreign co-authors who had never worked in Russia, against a real national researcher population of roughly three hundred thousand.

~ the case file ~

~ horticulture and pediatrics ~

The root error is easy to state. Web of Science and the Russian index assign subject categories to journals, not to individual papers. The Map took the journal's category, attached it to the author, and then aggregated authors into institutes. Nothing ever matched a person to their own publication list by identifier, so a paper could also be counted three times over, once as a Russian original in the domestic index and twice more as its translation in Web of Science and in Scopus.

What came out the other end was presented to the Presidium of the Academy of Sciences on 26 November 2013 and the room did not take it well. The Lebedev Physical Institute, one of the most decorated physics institutes in the world, had been assigned horticulture, pediatrics, gynecology, transport logistics and anthropology. Gazeta.ru ran it under the headline about physicist-gynecologists and mathematician-litterateurs. The Sobolev Institute of Mathematics in Novosibirsk had analytical chemistry as its first field and literary theory as its second, and under literary theory the Map ranked a mathematics institute third in the country, with a tomography center fourth.

Academician Ludvig Faddeev reported two theoretical physicists credited with patents on brucellosis treatment and on the butchering of meat carcasses. The Institute of the Earth's Crust in Irkutsk was listed with 327 scientists against an actual staff of about 115, the surplus consisting of the Taiwanese co-authors of a single paper, a Mongolian cosmonaut and the editor of an American geophysics journal, while the paper's own Irkutsk author was missing from it. Alexei Parshin's summary was that people simply laughed, and that he called it shoddy work and rudeness.

~ the map of Ivan Susanin ~

Two open statements landed in the same issue of Troitsky Variant in December 2013 and they gave opposite advice. The Council of the Society of Scientific Workers listed the founding mistakes, demanded regular progress reports, and objected categorically to the Map's data being used to judge any organization or any individual. It still told scientists to register and help test the thing.

The Public Control Commission on science reform went the other way. Its analysis was titled the Map of Ivan Susanin, after the guide who led an army into a forest on purpose, and it recommended that Academy institute staff boycott the system and submit no corrections at all. It had a point about the corrections: at the start, fixing an error in your own record required a paper form stamped by your employer.

~ what it costs to be counted wrong ~

This was a Russian state project from end to end, paid out of a federal program and argued about in the Presidium of the Academy, and I have never been inside any of that. The part that travels is the experience of being indexed. Anybody whose work gets counted has met a system that decided what their field was from where they published rather than what they wrote, and then watched a number derived from that decision get used for something real. The Map is what that failure looks like built once, nationally, on a four year contract, with no way to opt out.

~ address unknown ~

On 31 January 2017 the ministry's own Science Council issued a statement saying that after four years the tool had not reached any satisfactory level of quality, that a significant part of its information was wrong, and that it called on the ministry not to use the Map of Russian Science for any purpose in future. The following month the ministry's relevant department had it closed. The physical switch was thrown by the state agency on whose servers it lived. No decision document was ever produced, and the library that had been running it received no state assignment for user support that year.

The Audit Chamber's report of 29 June 2018 states flatly that the system has not functioned since February 2017, and adds the detail that finishes it: documentary confirmation of the Map's location was absent, and it sat on the balance sheet of neither the ministry nor the library. A frontend shell stayed reachable until about October 2018, its newest news item dated 7 November 2016. mapofscience.ru is a lot on a domain auction now, and the function passed quietly back to the citation index it had been built to replace.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »