Pick Topic
Review Topic
List Experts
Examine Expert
Save Expert
  Site Guide ··   
FAQ for "Missing Articles" HELP
The Major Question
Why does Expertscape seem to undercount my papers?

A brief, but incomplete, answer is:

  1. Expertscape puts papers into single-topic silos.
  2. For a given topic, Expertscape counts only those papers whose index terms include the topic as a "major heading."

More information is below.

What are all the reasons that Expertscape undercounts my papers?

First, it is important to understand that all author-pages in Expertscape apply to single topics. So, if you published 3 articles on asthma and 6 papers on pneumonia, then your asthma page will correctly show just 3 of your 9 total articles. An expanded discussion explains further.

Second, within single biomedical topics we've found that the best ranking of expertise requires Expertscape to omit certain classes of papers.

Putting it more positively, for each biomedical topic Expertscape includes only those papers:

Any paper meeting all of these criteria will be drawn into Expertscape's calculations without fail. We simply don't miss any.

The links above provide more information about selected criteria. In general, the "major heading" criterion is both the most unexpected one, and the one that most reduces an author's list of papers for any given topic.

But other search engines include all my papers. Why don't you?

Our mission differs from that of search engines. We calculate expertise. We are not an exhaustive guide to the literature -- PubMed does that fine already. As relates to our mission, we have found that considering non-major-heading papers adds nothing to the expertise signal, and in many cases detracts.

It may be reassuring to note that all authors are treated equally in this regard. In other words, all authors have only their major-heading papers counted, and all authors do not have their minor-heading papers counted.

Still, I think Expertscape is missing some of my papers. Who should I write to get them included?

You're really talking about correcting the PubMed records for your "missing" papers. Write to PubMed and/or the publisher of the journal that printed your article.

PubMed says to contact the journal publisher to correct basic citation errors such as:

  • author names
  • affiliations
  • bibliographic information such as date of publication, volume, issue, page, e-location
  • titles
  • abstracts

PubMed says to contact them to correct errors in their "value-added data on the citations," for example:

  • MeSH headings and subheadings (i.e. indexing)
  • Supplementary concepts
  • Publication types

This page is also relevant.

Of course, errata and retractions should be managed in the usual way.

Why can't I ask Expertscape to update my records?

Expertscape draws all of its information from PubMed records. To ensure that all authors are treated equally (an essential characteristic of our system), we cannot honor special-case requests.

Just as importantly, we have no way to honor such requests! In other words, our software is currently not configured to make author-level edits or additions to PubMed data.

In the future we may offer a way to link authors to papers for the purpose of correcting mistakes that Expertscape makes in such links. To be notified when that capability is available, please sign up for our low-volume email notification list.

It is unlikely that we will ever offer a simple way to alter PubMed's indexing of a paper. Because so much in the world depends on the PubMed database, it is in everyone's interests to correct errors in PubMed, rather than correcting them here.

Timing Issues
How far back does Expertscape look?

Expertscape looks back about 9 years.

The "about" qualification is needed because:

How often is Expertscape's database updated?

The Expertscape database is updated frequently, but not regularly.

That's all we'll commit to.

The date of the most recent update is always available here.

Why are my most recent articles not listed?

It can take a couple months for the PubMed team to index a paper after publication. Because Expertscape cannot use a paper without this indexing, it completely ignores all PubMed records lacking index terms. (Actually, it now appears that the indexing team is several months behind in indexing all but the highest-impact journals.)

Note that Pubmed may generate an incomplete, index-free record for a paper, sometimes calling this an "in process" record. Thus, your paper may have a record in PubMed and may even appear in some PubMed searches you perform, but the record will still be ignored by Expertscape because it has no index terms.

The solution is to wait for PubMed to finish their work (which they diligently do).

Why are some of my older articles not listed?

When a topic has a huge number of articles (for example, "autoimmune diseases," which has more than 130,000 in the past 10 years), Expertscape will silently drop some older articles that otherwise would have been included. It does this to keep computational demands tractable.

Currently, there is no good way for anyone to determine whether this has happened, but in practical terms it makes little difference, because (a) all article-authors are being treated equally, and (b) there are still a very large number of articles available to the system.

"Topics" and PubMed Indexing
What is "PubMed indexing"?

Indexing is the process of adding index terms to a paper's PubMed record.

This is not a simple task. Index terms are chosen from the MeSH vocabulary (circa 27,000 terms) and are, singly or in combination, assigned to a paper as major headings, minor headings, or modifiers.

A few journals do the indexing themselves, often using a vocabulary and syntax different from PubMed's. This is unfortunate.

What is a "major heading"?

PubMed's official definition of "major heading" is not that helpful, but, fortunately, the idea of an article having major concepts and minor concepts is easy to grasp.

So, in assigning MeSH terms to an article as part of the indexing process, the PubMed team will label some as "major headings," corresponding to the article's major concepts. As a hypothetical example, if an article about asthma mentions 12 different treatments, then asthma is likely to be one of the major headings for the article, and steroids, bronchodilators, etc. are likely not to be major headings.

As a result of this distinction, an asthma pharmacologist who wrote the 12-treatments article may expect that paper to count in his or her tally of articles about bronchodilators, but as far as Expertscape is concerned, it doesn't.

Why does Expertscape look only at major headings?

Quite simply, we have found that considering minor headings yields a noisier assessment of expertise.

My paper is indexed incorrectly. What should I do?

First, be sure your diagnosis is correct.

There are some subtleties in the use of index terms. For example, we have seen people misled by looking at the "nephrology" topic in Expertscape, expecting to see all of their publications related to kidney disease. However, PubMed includes in the "nephrology" topic only articles that are written about nephrology as a professional discipline, e.g. discussions of certification requirements, manpower, etc. The proper topic is "kidney diseases."

To forestall problems like this, you will occasionally see a special highlighted note on the overview page for nephrology and similar topics. Non-clinical topics are also called out, as most of Expertscape's use by laypersons is related to clinical topics.

Do you trust the PubMed indexing terms?

Yes. The PubMed team pays a great deal of attention to the indexing task. Nothing beats it.

The staff of the PubMed database makes the final decisions about the topics under which your articles are indexed. Expertscape has no input into that process and respects their decisions. If you can convince PubMed to make a change in their record, Expertscape will pick it up when they do.

Still, I think my paper is indexed incorrectly. What should I do?

The staff of the PubMed database makes the final decisions about the topics under which your articles are indexed. Expertscape has no input into that process and respects their decisions. If you can convince PubMed to make a change in their record, Expertscape will pick it up afterwards.

Assigning Papers to Authors
How does Expertscape assign papers to an author?

It's not easy.

Even a seemingly simple thing like an author's name can be confusing if:

  • It is erroneous, e.g. Robert Smith is published as Robert Smitth.
  • It is a variant, e.g. Robert Smith is published as Bob Smith.
  • It is changed or outdated, e.g. after a marriage or other event.
  • It is ambiguous, e.g. Robert Smith and Roger Smith both publish as R. Smith.
    • This problem affects common names most severely.
    • Especially in Asia.

Reliable, unique author-naming is a problem that bedevils all of scientific publishing. For the present, Expertscape has excellent algorithms to detect multiple names referring to the same person and then combine the names into one author identity. The algorithms are deliberately conservative, however.

When multiple names have been combined, it is so stated on the author's page.

What about ORCID?

PubMed records can include an author's ORCID, ISNI, or VIAF identifier. This will be important going forward, but is not yet widespread enough in the PubMed database for Expertscape to use.

The National Library of Medicine makes the point (5 minutes into this video) that journal publishers can retrospectively add an author's unique identifier to existing PubMed records. We encourage authors to contact journal publishers and take advantage of this.

My articles have not been published under a consistent name. Is that a problem?

It can be.

Any mis-assigned articles are still somewhere in the Expertscape system, but can often be hard to find. In the near future we may provide a mechanism to correct such mis-assignments. You can sign up to be notified when this capability is ready. (Sign up for Expertscape's low-volume notices for professionals.)