Skip to main content

ResearcherID, a unique identifier of a researcher

Submitted by Zhigang Suo on
h-index
researcherID
citation

On the Web of Science my name appears sometimes as Suo Z and sometimes as Suo ZG.  If I search for Suo Z*, papers by a biologist named Suo ZM mix in.  Now Suo is a very rare name.  I cannot imagine how Wang JS searches for his papers.  Last year Michelle Oyen and I talked about assigning a unique identifier to each researcher, much like assigning an ISBN to each edition of a book, or assigning a DOI to each paper.

This issue of misidentification is not lost on the good engineers at Thomson Scientific, the publisher of the Web of Science.  For months Thomson Scientific has been sending emails to promote a new service called ResearcherID.  I finally got around last evening and registered. 

Once you register, you are assigned a unique ID.  You can choose to make your list of publications public.  For example, here is my list:  http://www.researcherid.com/rid/B-1067-2008.  You can sort the list in three ways:  by title, by times cited, and by date added.  You can also view citation metrics such as the total number of articles in the list, the sum of times cited, and the h-index.

To obtain a ResearcherID, your institution must subscribe to the Web of Science.  Once you are on the Web of Science, you first register with the Web of Science, and then sign in to the Web of Science.  Subsequently, you register with ResearcherID.  Once you sign in to ResearcherID, you add publications to your account.

This morning I showed the procedure to John Hutchinson.  Here is John's list:   http://www.researcherid.com/rid/B-1221-2008.  We have noticed a few problems.  As expected, the Web of Science does not index all the papers published by a researcher.  Even for some of the indexed papers, ResearcherID does not seem to include all the papers in computing the citation metrics.  For example, John's list contains 247 papers, but the citation metrics are based on 108 papers.  ResearcherID shows his h-index to be 57, but if you go through the papers on the list, his h-index  is 68. There are many other aspects that you wish ResearcherID to do better, but this first release seems to be a good beginning.

ResearcherID may become a standard tool for a researcher.  Your administrators may ask for your ResearcherID, instead of a static list of publications.  You may want to include the ID in you resume and website.  (Your citation metrics are available to anyone who bothers to search.  Why not get the numbers correctly yourself and tell the world?)  Let's hope that we and our administrators can all think in many dimensions, with the ResearcherID being just one dimension.

Just for the fun of it, I have added a box in your iMechanica profile.  Once you have your ResearcherID, you can update your iMechanica profile.  Your ResearcherID will appear on the user list, and will also appear every time someone clicks on your photo.

Update on 18 April 2008. 

Yixiang Gan

I think it is extremly important to researches with Chinese names, since in chinese, the spelling and writing are not identical. This reminds me a note on Physical Review Letters PRL 99, 23001(2007), as following

Editorial: Which Wei Wang?

The
APS journals receive manuscripts from scientists all over the world.
For authors whose names cannot be expressed in Latin characters, their
names in the byline must be transliterated, a process that is not
necessarily bidirectionally unique. For example, the eight Chinese
names 王伟, 王薇, 王维, 王蔚, 汪卫, 汪玮, 汪威, and 汪巍 all transliterate as Wei Wang.
To remove some of the ambiguity arising from this unfortunate
degeneracy of names, APS will allow some authors the option to include
their names in their own language in parentheses after the
transliterated name, such as Wei Wang (汪卫). The option to present names
in the article byline in this manner is an experiment initially offered
to Chinese, Japanese, and Korean authors, whose names can be expressed
in Unicode characters. An example of a Japanese name is Tadanori
Minamisono (南園忠則), and a Korean name is Chang Kee Jung (정창기). In the
English text the given name precedes the family name, while the reverse
is true for the characters. As we gain experience, we may be able to
broaden this offer to other languages. Authors who wish to try this
option will need to prepare their manuscripts by following the special
instructions at

http://authors.aps.org/names.html.

Gene D. Sprouse

Editor-in-Chief

American Physical Society

Is it possible to include the original spelling in the mother language, especially for eastern asian languages?

BTW, a few decades ago, in China, we have different PinYin system, like Huang=Hwang。This makes someone difficult to find English papers writen by Prof. Hwang Kehchih.

Wed, 04/16/2008 - 14:56 Permalink
Zhigang Suo

The method proposed by the APS editor is unlikely to solve the problem of misidentification.  Two people can have the same name with identical characters.  Besides, I don't read Karean and many others don't read Chinese. 

To solve the problem of misidentification, most likely we'll have to assign each author a unique identifier.   We can try to have our ResearcherIDs included in our papers submitted to journals, printed along with our addresses.  If many researchers do this, it is likely that publishers will ask for ResearcherIDs, just as they ask for email addresses.

The main difficulty seems to be this:  which ID should prevail?  The ResearcherID offered by Thomson Scientific, or some other ID created by another entity?  In order for the ResearcherID to be universal, Thomson Scientific should make the service more open, allowing people to register with ResearcherID without subscribing to the Web of Knowledge.  Coordination between ISI and publishers seems to be needed.  If no agreement can be reached, a unified system of IDs may be required by the government, or offered by companies like Google or Amazon.

Fri, 04/18/2008 - 12:31 Permalink
Zhigang Suo

Several people have asked how to register for ResearcherID.  Here was how I registered.  (Update 18 April 2008:  Teng Li has just pointed to a visual instruction.)

  • Go to ISI Web of Knowledge through your institution, as if you were searching for a paper.
  • On the right side, register with the Web of Knowledge.
  • Sign in with the Web of Knowledge.
  • On the right side, register with ResearcherID
  • Click Add Publications, and you can begin to add papers to your account.
  • Edit profile of your ResearcherID.  For example, you can make your citation metrics public, saving other people from doing arithmetics.

You can also add your ResearcherID to your iMechanica profile.  Your ResearcherID will appear on the user list, and will also appear every time someone clicks your photo.  You now have one more reason to be an active user of iMechanica:  by being active, iMechanica keeps advertising your papers to other mechanicians.

Let us know if you find an easier way to register. 

Thu, 04/17/2008 - 00:54 Permalink
Kilho Eom

It is very easy to register Researcher ID and also easy to include the

publications to Researcher ID. It seems very useful to manage my

publication list for my own purpose as well as to provide the appropriate

publication list to one who is interested in the researches conducted by

a personnel registered to Researcher ID.



I will also update my iMechanica profile to include the Researcher ID.

It is quite simple to use Researcher ID!!!



Kilho

Thu, 04/17/2008 - 08:07 Permalink
MichelleLOyen

Interesting comparison, I had never used ISI before this experiment with Researcher ID and when I make a direct comparison the numbers are far "better" for me in Scopus: compared with ISI, I have 50% more publications (although this is in part because ASME and MRS proceedings are included), 35% more times cited (interesting since it's not like the proceedings are actually cited that often so that does not account for the difference) and an h-factor 2 higher.  It will be interesting to see how this all shakes down.

Mon, 04/21/2008 - 11:38 Permalink
Teng Li

First of all, this is great service, with many things to be improved.

I have no difficulty to build up publication list, but share Michelle's difficulty in setting up citation metrics. I guess it should be done automatically but it keeps saying:

"None of the items in your publication list were added using the

Web of Science options on the Add Publications page."

Another observation is that, the publication titles are hyperlinked in my firefox browser, but not in IE. 

Thu, 04/17/2008 - 21:11 Permalink
jeroen bosman

Hi all,

 

here in utrecht (The Netherlands) we subscribe to both Scopus and Web of Science. Scopus indeed seems to have broader coverage in the technical sciences, as it includes the full Compendex database. We also did a comparison of copus, Web of science and Google Scholar.

For over two years now Scopus has had author identification, with unique ID's for each researcher. The system automatically groups name variants and shows a profile of each researcher included the metrics (H-index etc.). It assigns the articles based on author name, subject area, affiliation history and citation paptterns. Of course some article are left unassigned and very very rarely an article will be wrongly assigned. That's why authors can give feedback and suggest grouping or ungrouping of articles and author name variants.

E.g. the author ID of Michelle Oyen is 6602216219 (30 grouped articles), that of John W. Hutchinson is 7402097869 (223 grouped articles). It is nice that Scopus attempts this, although user feedback will prove essential to get all articles correctly assigned. It will be interesting to see how these two systems of author ID's develop.

Wed, 04/23/2008 - 12:28 Permalink
MichelleLOyen

I made the point previously that this ID in Scopus in incomplete for me.  If I search for "Oyen M" and then check box the names that are actually really me, there are six boxes to check and a total of 46 publications.  The Author ID is thus only catching 2/3 of my work and there does not seem to be any mechanism for feeding that information back to Scopus, at least not that I could find.

Wed, 04/23/2008 - 20:11 Permalink