Who wrote that?

Much of mathematics is identified with people.  We don’t talk about Algebra, we talk about “Lang”.  We speak of “Feller”, not An Introduction to Probability Theory and Its Applications.  But which Feller is actually this Feller? Here are some choices:

The answer is the last Feller in the list, William Feller.  Being able to tell one mathematician from another is important, especially if it is someone whose work you would like to follow.  Or if it is your own work.  From the very beginning of Mathematical Reviews (in 1940), we have made a point of distinguishing authors, what is known as author disambiguation.

I have a few favorite examples of mathematical names that need some work to be distinguished.  The first is my Steve Miller Problem.  I once needed to send a manuscript on number theory to Steve Miller.   Shortly after the manuscript left my hands, I had an email message, “You have the wrong Steve Miller.”
“Aren’t you Steve Miller the number theorist?”
“Yes, but I’m not the one you want.”
“Didn’t you do your PhD with Peter Sarnak?”
“Yes, but I’m not the one you want. ”
“???”
Checking MathSciNet + the Mathematical Genealogy Project shows that there are indeed two Steve Millers meeting all those criteria.  We have Steven J. Miller, my “wrong” Steve Miller, and Stephen D. Miller, my “right” Steve Miller.  Indeed, searching for “Miller, Ste*” results in eleven matches:

But it could be worse.  Suppose you were looking for Li Li.  There are 87 authors with that name, 30 of which we have verified.

Screen Shot Author Li Li 1

The names with the superscripts are those that we have verified.  If you want to zero in on a particular name, you get some help sorting through them by hovering your mouse over the Hover icon to see some information about a sample publication:

Screen Shot Author Li Li hover

So how do we verify an author?  This is a job for our Cataloging Department!   The idea is to match the name with an author already in the database.  If we have 80 instances of the same name (and this is not the record – there are 162 authors named Wei Wang, for instance) with 30 verified, how do we do that?  Most articles arrive with extra information, such as an email address and the name of the institution (including department), both of which are very helpful in identifying the author.   We also look for coauthor matches.  If we have a paper from Li Li that is coauthored with Fernando Paganini, this points us to Li Li9. Self references (i.e., if you cite yourself in the references) are very useful, too, as are keyword matches in the title.

If none of these proves helpful, we will then go to the internet. If the author has provided an email address on the article, we can use that to find a website. Otherwise, we will search for the author’s name and institution, and hope to find a website. We then use the author’s website to try to find evidence of a match, such as the posting of a preprint of the article or the mention of research on the topic.  If this fails, we will send the author an email message asking “Are you the same person who also wrote the paper ….” and give the title and bibliographic information of another article we have in our database.

If there is no email provided, we will make a note of it and hope that the next time they publish, they will use an email address.  In the good old days, we would send a letter, which took a long time and didn’t always work.  Now it works so infrequently that we no longer send letters.

At the time of this writing, we have indexed 767,027 authors.  So rare events happen.  For instance, Angelo Plastino and Angel Ricardo Plastino often list the same institution (and department) in their publications and write about similar topics! And both have coauthored papers with Flavia Pennini.  Angel usually publishes with an initial, which is helpful, but Angelo usually just publishes with A. Plastino. Many emails have been sent to confirm which papers belong with which, but luckily A. Plastino always responds with the answer!  Going the other way, the same person can be known by many names.  For instance, Yuriĭ Alekseevich Mitropolʹskiĭ (Юрий Алексеевич Митропольский) has over 40 variations on his name!

One of my themes is that Mathematical Reviews makes a great effort to ensure that the information in MathSciNet is right. (Look here or here.)  In this post, I have described one way that the experts in our Cataloging Department put their hard work and talents to use to achieve that.  Fortunately for users of MathSciNet, there are great people in every department.

Thank you to Elizabeth Downie for her help with this post.

About Edward Dunne

I am the Executive Editor of Mathematical Reviews. Previously, I was an editor for the AMS Book Program for 17 years. Before working for the AMS, I had an academic career working at Rice University, Oxford University, and Oklahoma State University. In 1990-91, I worked for Springer-Verlag in Heidelberg. My Ph.D. is from Harvard. I received a world-class liberal arts education as an undergraduate at Santa Clara University.
This entry was posted in General information and tagged , , . Bookmark the permalink.

One Response to Who wrote that?

  1. Jerry Grossman says:

    A corollary of this article is that every mathematician should look at what Mathematical Reviews (MathSciNet) thinks is their list of publications and let MR know about any discrepancies (in either direction—your papers that MR thinks belong to someone else, or papers that MR has erroneously attributed to you). In the process of developing and updating the Erdos Number Project website (which lists all people with Erdos number less than or equal to 2), I have come across some interesting instances of author ambiguity (some of which will forever remain unresolved) and have occasionally been able to help MR with their noble goal. There are many bibliographic websites out there, but none compares to what MR has to offer in terms of treating authors as people and not name-strings. Keep up the good work!

Leave a Reply

Your email address will not be published. Required fields are marked *

HTML tags are not allowed.

13,321 Spambots Blocked by Simple Comments