Ticket of the month - October 2023 - Dispelling pesky journal-title-level registration errors

In the past year, our technical support team has answered more than 4500 tickets related to content registration. One of the more common content registration problems or points of confusion is with our title check/validation process. For this post, I am going to focus on journal-title-level content registration issues since that’s the most common record type registered with us. But, we also have a rigid title check/validation process in place for other record types, too (e.g., books, conferences, etc). And, this rigid title check/validation process is only at the journal-, book-, conference-title-level (or, container level); so, for instance, we do not have a similar check/validation process in place for journal articles, book chapters, conference papers, etc.

Some process-related background first
When you, as a member, send us XML - either by HTTPS POST, direct upload to doi.crossref.org, or in using our helper tools, like the web deposit form, to help build that XML for you - we store the XML you’ve uploaded and process it asynchronously from your submission.

If you’re registering using a direct upload or web deposit form, we give you a message to say that we’ve received your file successfully, but there could be problems later once we start actually processing it. Again, think asynchronously. There’s a lag in the process.

So you could see a message like this:

Or, this if using the web deposit form for your registrations:

But still have submission failures that prevent your DOIs from being registered. That success message only means that we have successfully received your submission(s) and are awaiting a free submission processing thread to register your metadata with us. It’s usually a fairly quick process (minutes), but there can be lags (e.g., many other members registering metadata at the same time as you or during times of Crossref maintenance).

Once we have completed processing your file, we will send you another message to tell you whether our processing has been successful. That’s why it is important to pay close attention to the emails we send you as a result of your submission(s). Those emails contain submission logs with the results of your registrations. You should be looking for success messages in those emails, like this:

When you see that ‘successfully added’ status in your submission log, it’s time to celebrate. In the screenshot above, DOI 10.1348/000709907X203706 has been successfully registered. The DOI is active. Once you see this confirmation, you may resolve the DOI link https://0-doi-org.pugwash.lib.warwick.ac.uk/10.1348/000709907X203706. From there, we’ll be working on indexing the full metadata record in our APIs and in search.crossref.org. There are lags in that process (some up to 24 hours), but, if you see these successes, then we’re hard at work on next steps in this process.

[Note: in this example, there were nine total DOIs in the XML file (i.e., 9 record count), and all nine of those DOIs were registered (i.e., 9 success count).]

Journal title checks
As some of you may know, our title check process is rigid. It expects and, in order to run smoothly, relies on consistency. It was built rigidly so that the metadata at those container levels (e.g., journal title) registered with us is as consistent as it can be. But, that rigidity leads to many failures - as is evident with over 4500 tickets submitted to our technical support team on the topic in the last year.

When you send us the initial registration for your journal, we establish a journal title record in our system that we use to compare against for subsequent registrations. For this post, let’s use an example to help. Here are my example journal details:

Journal title: The Journal of Stuff, Things, and Goodies
Abbreviation: JSTG
Journal-title-level DOI (or, journal DOI): 10.5555/tm34ej0
The landing page for the journal as a whole (or, journal URL): https://0-www-crossref-org.pugwash.lib.warwick.ac.uk/blog
Electronic ISSN: 98764321

And, here is how I have entered that metadata into the web deposit form for my initial registration of this journal:

As a result of this web deposit form submission, we create a title record within our database, visible to our technical support team in the backend of the admin tool, that looks like this:

So, some things to observe here:

  1. We’ve normalized (out) the commas in the title record. This gives our title check process some much-needed flexibility. Because of this, if I resubmit the title without using the Oxford comma - The Journal of Stuff, Things and Goodies - our system will not reject this subsequent submission. We normalize commas, hyphens, apostrophes, periods, and colons in our title records.
  2. The title, eISSN, and DOI are set. We do not expect these values to change in subsequent submissions. If they do, you can expect failures for those subsequent submissions. [Note: if you have a print ISSN and electronic ISSN, include both in your registrations for DOIs published in both the print and electronic versions of your journal; if you’re inconsistent with the ISSNs you submit, that can lead to multiple title records in our system that may need our intervention to merge.]
  3. This journal is owned/associated with one DOI prefix - 10.5555, the Crossref Test account. Only new DOIs on prefix 10.5555 can be registered for this journal.

Okay, let’s talk errors
People and their systems make mistakes; we get it. We’re here to help if/when that happens. But knowing the information above should help in registration planning and execution.

Now that we have established a baseline of how this title check/validation process works, let’s discuss what happens if/when we stray from consistency and encounter errors.

Hey, the title’s wrong! Please help!
Journal-title-level inconsistencies or errors are typically realized at the point of a subsequent registration. Maybe a member notices the error message in the submission log we email them. Or, maybe DOIs they thought they had registered are not resolving. When members raise these issues with our technical support team, the first thing we do is investigate for discrepancies. We ask ourselves: what’s inconsistent and not matching between the current registration and the initial or previous registrations?

I made a mistake when I initially registered my journal title
Let’s say that in my example of The Journal of Stuff, Things, and Goodies, I just completely missed the memo about which title was the official title registered with the ISSN Centre and made a mistake when I registered the title with Crossref as The Journal of Stuff, Things, and Goodies. As a result, my team flags the error to me:

Isaac, the official title for the new stuff journal contains an ampersand [i.e., The Journal of Stuff, Things, & Goodies; something looks off with my registrations when I include that title with the ampersand in my registrations for the most recent issue! What happened?

Maybe one of my colleagues is responsible for these subsequent registrations and is receiving an error like this from our system (again, they’re seeing this in the emailed submission logs):

<record_diagnostic status="Failure" msg_id="22">
<doi>all doi's under the current journal element</doi>
<msg>ISSN "98764321" has already been assigned, issn (98764321) is assigned to another title (The Journal of Stuff Things and Goodies)</msg>

What? Why?

Well, as we discussed, the title for the journal as a whole was previously registered as: The Journal of Stuff, Things, and Goodies and this subsequent submission includes the title for the journal as a whole as: The Journal of Stuff, Things, & Goodies. That small discrepancy between the use of the word and and the & character has caused all of the DOIs in this submission 1000600009 to fail. None of my DOIs in this submission have been registered.

In order to resolve this issue, there must be consistency in the journal titles submitted. So, we have two options.

Option 1: Update the journal title in the recent submission to match what we have in the title record. So in our example, update the journal title in the submission to: The Journal of Stuff, Things, and Goodies.

Option 2: Contact our technical support team to update the title for the journal as a whole to: The Journal of Stuff, Things, & Goodies. If we did that, all subsequent submissions would have to include the version of the title with the ampersand.

Here’s what the XML for these two titles would look like:

<full_title>The Journal of Stuff, Things, and Goodies</full_title>


<full_title>The Journal of Stuff, Things, & Goodies</full_title>

This is the comparison our technical support team is performing when one of these journal-title-related registration errors is reported to us.

What’s with this journal DOI thing?
Registration of journal-article-level DOIs is mandatory. These journal-article-level DOIs are really unique and persistent citation identifiers.

But there is also the option to register DOI records for the journal title itself, or even a particular issue or volume.

Screenshot 2023-11-02 at 4.20.01 PM

Under a traditional, let’s say, quarterly journal publication model, our members would organize their journal registrations into annual volumes that contain four issues. In each issue, that member would publish and register a number of journal article DOIs.

Again, for the sake of illustration, let’s say that The Journal of Stuff, Things, & Goodies has been publishing for ten years. We publish four issues per year in a volume and each issue contains ten journal articles. I can register DOIs for all 400 journal articles, all 40 journal issues, all 10 volumes, and one DOI that identifies the journal as a whole. Journal-issue-, journal-volume-, and journal-title-level DOIs are all optional. The greatest utility from a citation identifier perspective is for the journal articles to each have DOIs, since journal articles are often cited, shared, and recorded. Journal-issue-, journal-volume-, and journal-title-level DOIs are almost never cited, but they can be shared and recorded, so they are important, but not as integral as DOIs for journal articles.

We have many members who don’t use a traditional quarterly journal publication model, but want to include rich metadata for the DOIs they register with us. We appreciate that. If these concepts of journal-issue-, journal-volume-, and journal-title-level DOIs are new to you, and you don’t quite understand them, omit them from your registrations.

If you want to register DOIs for journal issues, volumes, and your journal as a whole, be consistent with the DOIs you’re registering. If you register a journal-title-level DOI, record it, share it with your colleagues, and be sure that you think of it as THE IDENTIFIER for the journal as a whole. You’ll want to do the same for your journal-issue- and journal-volume-level DOIs, but the journal-title-level DOIs are checked during the title check process/validation and inconsistent use of journal-title-level DOIs can create registration errors.

In my example, I’ve previously registered my journal-title-level DOI as 10.5555/tm34ej0. If I forget that DOI or fail to share it with my colleagues, and then during the next quarter we’re registering DOIs and use a different journal-title-level DOI, we’re going to encounter errors. Let’s say my colleague enters into the web deposit form:

Unfortunately, it appears as though they’ve conflated the idea of a journal-title-level DOI with a journal-issue-level DOI. But, whatever the cause, their submission, since I previously registered the journal-title-level DOI as 10.5555/tm34ej0, will result in this submission error:

<record_diagnostic status="Failure">
<msg>Deposit contains title error: The journal has a different DOI assigned; If you want to change the journals DOI please contact Crossref support: title={0}; current-doi={1}; deposited-doi={2}</msg>

Now, like with the title errors above, we’re back to asking ourselves which of the two DOIs will be the persistent identifier for the journal going forward? And, we need Crossref support to provide guidance (which we’re here for).

My journal’s title has changed [substantially]
The ISSN Centre does permit minor changes of journal titles without an ISSN change. For instance, updating my official journal title from The Journal of Stuff, Things, and Goodies to Journal of Stuff, Things, & Goodies would not necessitate new ISSNs for the journal.

In cases of substantial or significant journal title updates, the ISSN Centre will recommend new ISSNs. And, that’s a very good thing, since it will mean your citations between the old journal title/ISSN and the new journal title/ISSN are clean and not conflated by our citation matching process. For instance, if I wanted to update my journal title from The Journal of Stuff, Things, and Goodies to Goodies International. In this case, the ISSN Centre would recommend a new ISSN for the journal using this significantly different title. Our system, again for the reasons of clean citation matching, would effectively treat the updated journal title as a new journal with a new journal title and a new ISSN(s). All the existing DOIs registered using the previous title The Journal of Stuff, Things, and Goodies should remain associated with the previous journal title and ISSN, while new registrations of DOIs would adopt the new journal title Goodies International and new ISSN(s) from the ISSN Centre.

For substantial or significant journal title changes, where a new ISSN(s) is recommended by the ISSN Centre and used in your registrations, your initial registration against this new journal title and new ISSN(s) will create a new journal title record in our system that we will use to check/validate your future new journal registrations against.

We’ve read a lot; let’s wrap this thing up
Consistency with the metadata being submitted at the journal-title level is critical. When submitting your initial registration for a journal, be careful that the journal title, journal-title-level DOI (if you have one), and ISSNs are all entered correctly and that information is shared widely at your organization so it can be recorded and used consistently by your colleagues in future registrations. Sometimes mistakes happen or things change over time, and we’re here to help when they do, but investing extra care when entering and registering journal-title-level metadata with us can minimize headaches in the future.

Thanks for reading,