Some thoughts on unique author IDs

- 13 Comments

What I want to write about today is this idea, seemingly fashionable
on blogs and FriendFeed (and PLoS editorials) at the moment, of
authors switching from using their names to identify themselves on
papers to using some other form of ID. The so-called unique author ID
problem.

Now if your name is Xanthe Unique, you might be wondering what
the problem is exactly – you just write your name and that’s it. It is
indexed in Scopus/ISI/whatever subject-specific repository holds sway
in your field. However, there is a problem where you may not being
credited for your work if any of the following conditions apply:

1 – You share your name with someone else. There is a unusual case in high
energy physics where there are two people called Michel Tytgat working in approximately the same area. Obviously more likely if you
are called John Smith or Jesus Garcia.
2 – Your name is
transliterated
. Maybe you are Chinese, Japanese, Korean, Thai, or any
of the places which do not use the Roman character set. In order to be
included in the relevant database/journal/website etc your name needs
to be transliterated to Wei He or whatever. Sometimes it could be
transliterated slightly differently (Wei Hei) and sometimes this
can overlap with shared name problem described above in the western character set where it
never existed in the original language.
3 – You have changed your
name
. Most common case here is the women who get married and take their
husband’s names. Jayne Smith changes to Jayne Brown, but is still the
same person.
4 – You write your name differently. In order to
identify yourself correctly each time, you need to write your name the
same each time on each paper, or make sure your coauthors do. It could
be the case that the same person is identified as (taking myself as an
example) – C. Leonard, C.J. Leonard, Chris Leonard, Christopher Leonard, Christopher J. Leonard, Christopher James Leonard etc.

So
how to make sure that your work (and citations) are being credited to
you? Well, one idea starting to gain sway is the concept of not
referring to yourself by name, but by some other unique identifier such
as a number.

The advantages are clear. All of the confusions
above could be circumvented and any aliases could be documented on a
unique author homepage.

So how might this happen? Well, this is
something I spent some time thinking about around a year ago at a
discussion about the next generation of bibliographic services in
high-energy physics. I was on a mini team looking at the unique author ID problem and whilst this team was looking at HEP in particular, I was
aware that this was an opportunity to solve the problem for all of
science (note that Scopus has already implemented something like this, all be it not open and behind a subscription wall).

Here, with some image mockups from the time, are some of my thought on the matter.

1 – A central, independent website is in charge of assigning and cataloging all author id numbers.
2 – An author comes to the website to be assigned a unique author id.
3 – The author uses this id in addition to his name on all scientific papers (and publishers publish them alongside the names)
4
- The author has a homepage on this central repository website where he
can list aliases, former institutions, and we can autogenerate a
bibliiogrpahy (which in itself brings up issues about a&i databases
in future). The author can also link to other identifiers about himself
such as OpenID etc. Photos, links to blogs, homepages etc could also be
inserted here. Could also list old email addresses which no longer work.
5 – Anyone can apply for an author id. Authors of grey literature are also encouraged to use this UAID.
6
- This website will not try to validate the applicant as being unique.
It is in their own interests to apply once and use only one UAID number.

Who
would operate this website? A cross-publisher body such as CrossRef
would seem to be one candidate, or otherwise it would be a wholly
independent body – possibly funded by publishers. Clearly I need to
work on it a bit more and I am almost certain it has overlap with many
of the other discussions on the web at the moment, but if I read them
all, I wouldn’t be able to get any work done!

I have also decided to share my slides from the time on SlideShare.

Unique Author IDs
View SlideShare presentation or Upload your own. (tags: uaid)

Let me know what you think.

  • http://alexholcombe.wordpress.com/ alex holcombe

    good ideas!

  • http://names.mimas.ac.uk/ amanda hill

    Hello Chris

    The JISC has funded a prototype project to look at creating the sort of service that you describe. The Names Project is a collaboration between Mimas (at the University of Manchester) and the British Library.

    We’re coming to the end of the prototype development phase and hope to develop a pilot service over the next two years.

    Amanda

  • http://www.physmathcentral.com chris leonard

    Hi Amanda,

    Thanks for that info. I wasn’t aware of the Names project, but it looks interesting. However I think there is a pressing need for a simple service right now and that by the time 2 years has passed there is a danger some other service will have scracthed this particular itch already.

    Some peope on FF seem to think it is a question of connecting a few web services together and ensuring permanence as long as possible – however I’ve seen no mention of the Names project yet so will alert everyone to this.

    Cheers,
    Chris

  • http://bjoern.brembs.net björn brembs

    http://getcited.org has similar functionality but is dead (probably due to lack of automatization).

  • http://names.mimas.ac.uk/ amanda hill

    Sorry, wasn’t clear enough – we’ve got the prototype functioning now (with a very small set of test data) and are planning to further develop it into a pilot service (in an iterative way, responding to feedback from those who need such a service) in the next two years (assuming that the proposal is funded – can’t really promise anything until that decision is made).

  • susan besaw

    Hello Chris,

    You may be interested in taking a look at ResearcherID- http://www.researcherid.com. This free website assigns members a unique identifier and provides the ability for published authors to build a profile for promotion of their publications and research interest. Visitors to the site can search the registry to find collaborators, review publication lists and explore how research is used around the world.

    You can automatically request to register via this

  • http://www.physmathcentral.com chris leonard

    Hi Susan,

    Yes, researcherid has many components of what I was thinking. Looks very interesting – I will have a good look round the site. How long has it been going? I’m surprised and a little disappointed I hadn’t heard of it yet.

    Chris

  • http://www.slac.stanford.edu/spires/ travis brooks

    Hi Chris,

    Since I was organizing that summit of HEP information providers, I remember very well what got you thinking about this knotty issue.

    The key thing that I remember from the time is that we needed a 3rd party, central site that covered an entire discipline, like SPIRES (or INSPIRE, which grew out of those discussions). What made us defer the author ID project at SPIRES was the need to give people an easy way to claim their papers. INSPIRE will take over from SPIRES in the Fall of this year, and we hope to, soon after, allow authors to claim papers.

    Our solution will work well in HEP, where researchers and author communities overlap, and where our 3rd party site is already in heavy use. The problem is creating the uniformity and central authority in disciplines that don’t already have them. Glad to see you’re still thinking about that.

    Best,
    Travis

  • http://www.physmathcentral.com chris leonard

    Hi Travis,

    Yes, it’s something that has been occupying my mind since that INSPIRE kick-off meeting.

    It is clear that all parties are on top of what is needed for the HEP community and are working towards that goal – a somewhat unique situation in the grander STM world where this kind of collabroation and cooperation isn’t typical.

    It would be nice if INSPIRE were to cover all of science, or of if other disciplines were to copy INSPIRE’s lead. In the absence of this, some other central, open repository of all author ids for all of science would seem to be the way forward.

    Note that this does not exclude enhanced offerings being made in certain disciplines (INPSIRE), it’s just that these should link to this central repository as well. I think this is particularly helpful for interdisciplinary areas of research where a researcher in e.g. medical physics, may be equally interested in HEP as biology.

    In any case, it is clear that CrossRef are about to offer something along these lines soon (CrossReg) and researcherid (above) already covers much of what I blogged about.

  • etienne joly

    One major problem with setting up UAIDs seems to be the identification of a single provider of these IDs, and the monopoly that would result from it. So I feel like asking a provocative question: does one really need to have only one UAID provider ? When nucleotide databases were started, new sequences were communicated either to EMBL or Genbank, or even to other, more specialised, repositories. Pretty soon afterwards, the databases realised that they had to share all of their contents, and sequences ended up with two (or more ) accession numbers. But if we look back in time a little bit further, we will see that, at first, scientists could publish their new sequences without communicating them to databases, and they quite often did so, either out of laziness, or to preserve the privilege of having the computerised form of those sequences. It is only when most publishers started DEMANDING that authors provide the accession numbers of those sequences that nucleotide databases became really exhaustive and useful. I think that UAIDs will only become really widespread if most publishers start requesting that all authors on submitted papers provide at least one type of unique author ID ( as well as funding bodies from applicants for money). And it would not have to be with one single UAID operator. Different UAID providers could offer different types of interface and/or services ( attached web pages, blogs, bibliometry analyses …) , as long as the minimal info was provided ( ie the list of publications vetted by the authors themselves, in an open access format). The question will then evolve into the openness and interoperability of the various systems, but that is something that should be much easier to solve than sorting out between authors with identical names !

  • chris leonard

    Etienne –

    Absolutely I agree, and services such as http://sameas.org/ facilitate this. The onus may well be on publishers to insist on some form of ID which can then be queried in a database of IDs to resolve to the correct individual.

    Chris

  • http://efoundations.typepad.com/ andy powell

    You seem to have ignored or ruled out a user-centric solution such as that offered by OpenID? I.e. that the solution lies in each user having one (or more) ‘http’ URIs for themselves and you leave it up to them to find an appropriate host for that/those URI(s). Host service providers might include universities, repositories such as arXiv, people’s own websites, other existing OpenID providers, Google, Facebook and so on.

    Allowing people to adopt their own identifiers is a much more Web-like approach – as opposed to giving some centralised identifier-issuing service a monopoly position in the space.

    On the other hand, clearly, there are some short term implementation/take up benefits to a centralised solution.

    So… two things (in order of importance).

    1) Any solution must be based on ‘http’ URIs (URLs if you prefer) because that is what the Web is built on).

    2) A user-centric solution is better (in principle) than a centralised solution (but that statement trivialises a rather complex set of issues).

  • chris leonard

    Andy –

    I am sure you are right – and, so as not to rule other services in or out, I think the way forward could be to use something like sameas.org to associate various online ids (including OpenIDs or whatever people want to use) into one aggregated – rather than centralised – resource.

    http://sameas.org/about.php

    Chris