TL;DR We missed an error that led to resource resolution URLs of some 500,000+ records to be incorrectly updated. We have reverted the incorrect resolution URLs affected by this problem. And, we’re putting in place checks and changes in our processes to ensure this does not happen again.
How we got here Our technical support team was contacted in late June by Wiley about updating resolution URLs for their content. It’s a common request of our technical support team, one meant to make the URL update process more efficient, but this was a particularly large request.
Crossref Conversations is an audio blog we’re trying out that will cover various topics important to our community. This conversation is between colleagues Anna Tolwinska and Rosa Morais Clark, discussing how we can make research happen faster, with fewer hurdles, and how Crossref can help. Our members have been asking us how Crossref can support open science, and we have a few insights to share. So we invite you to have a listen.
We’ve just added to our input schema the ability to include affiliation information using ROR identifiers. Members who register content using XML can now include ROR IDs, and we’ll add the capability to our manual content registration form, participation reports, and metadata retrieval APIs in the near future. And we are inviting members to a Crossref/ROR webinar on 29th September at 3pm UTC.
The background We’ve been working on the Research Organization Registry (ROR) as a community initiative for the last few years.
We’re excited (and a little nervous) to launch a new research project designed to assess the effects of metadata on research communications. We’re expecting this effort to be a significant contribution to the existing research on the topic and we’re really looking forward to getting started. We’re also a little nervous because of course we don’t know what the conclusions will be (after all, if we did, we wouldn’t be starting this project).
Of these three parts of the DOI, members (or their service providers) create the last part, the suffix. Because DOIs must be unique and persistent, members need a reasonable way to create and manage their suffixes, which should be opaque.
Here, we share the rules, guidelines and some examples to help you decide how to approach your suffixes. You can also go straight to our suffix generator.
Suffixes are case insensitive, so 10.1006/abc is the same in the system as 10.1006/ABC. Note that using lowercase is better for accessibility.
Guidelines for creating a DOI suffix
In part because there are few rules, it can be helpful to have some guidance in how to approach suffixes. This advice applies to DOIs at all levels, whether at journal or book level (a title-level DOI), or volume, issue, article, or chapter level.
The most important part of creating your DOIs is to understand that because DOIs are unique, persistent and ‘dumb', once they are created, they will always work. There is never a need to delete or update existing DOIs.
Best practices for DOI suffixes:
Suffixes are best when they include short strings that are easily displayed and typed but are ‘dumb’ - meaning, the suffixes contains no readable information, including metadata.
Keep suffixes short. This makes them easier to read and to re-type. Remember, DOIs will appear online and in print.
Best practice DOI example:10.3390/s18020479 This example appears to be opaque because it includes no obvious information.
Avoid the following in DOI suffixes:
The function of suffixes is technical in nature so they are most problematic when they are treated as information to be read, interpreted and/or predicted. Remember, DOIs are persistent and not subject to correction or deletion.
While it may be tempting, using a pattern, such as a sequence, can cause problems. Services and tools that use DOIs may, for example, try to predict future DOIs that are not registered and may never be (more on opaque suffixes below).
Don’t include information like journal title (or initials), page number or date. This kind of information should be included in the metadata but can cause problems when included in suffixes for 2 main reasons:
Information in the suffix that conflicts with information in the metadata is confusing.
Information like journal title (or initials) may change or be found to be incorrect, as with dates, but DOIs are persistent, cannot be deleted and are not subject to correction. See more on opaque suffixes below.
Example problematic DOI suffix:10.5555/2014-04-01 This example is not opaque because it includes a date, which should be included in the metadata instead of in the suffix.
Proceed with caution in DOI suffixes:
Determining how to create suffixes and manage the over time can be a challenge. We recognize that some systems have requirements that don’t follow this advice and that human readability is helpful in managing DOIs.
If you must use a suffix with meaning, internal system identifiers can work, with careful management. Because things like ISBNs are themselves metadata, we don’t recommend using them in suffixes.
Just remember that while you and readers may recognize an ISBN, for example, the DOI system itself doesn’t and DOIs are not subject to correction or deletion.
No matter your approach, it’s worth taking some time to understand the emphasis on opaque suffixes.
What if your content already has a DOI?
Sometimes members may acquire a journal that already has DOIs registered for some articles. It’s important to keep and continue to use the DOIs that have already been registered and not change them - DOIs need to be persistent.
It doesn’t matter if the prefix on the existing DOI is different from the prefix belonging to the acquiring member. As content can move between members, the owner of a DOI is not necessarily the same as the owner of the prefix. Read more about transferring responsibility for DOIs.
The importance of opaque identifiers
What are opaque suffixes & why they are important
Suffixes are ‘dumb numbers.’ They are essentially meaningless on their own and meant to be that way–opaque. One good reason for that is because when something is meaningless, it doesn’t need to be corrected.
DOIs should not include information that can be understood, interpreted or predicted, especially information that may change. Page numbers and dates are examples of information that shouldn’t be included in suffixes. It is particularly problematic if the suffix includes information that conflicts with the metadata associated with the DOI.
We’ve referred to creating ‘suffix patterns’ in the past but information that includes or implies a pattern is also problematic. A sequence of numbers, for example, lends itself to the assumption that future DOIs can be predicted.
Scraping for DOIs - or what appear to be DOIs–is common, as is the likelihood that what is–or appears to be–a pattern will be treated as such. Just as the timing of DOI registration is important, in order to avoid unregistered DOIs, their construction is critical to avoiding interpretation.
More information on creating DOIs
Here are a few other resources that discuss creating DOIs and the importance of using opaque suffixes.