Collecting and Analyzing Scholarly Metadata from Scratch with the Crossref REST API (for absolute beginners) (Part 1)

What’s Crossref?

Crossref is a not-for-profit membership organization that connects the global scholarly research community by collecting, enhancing, storing, and freely sharing the metadata deposited by over 24,000 research-work-producing member organizations (academic publishers, universities, funders, museums, libraries, hospitals, etc).

It’s the organization that’s responsible behind-the-scenes for 95% of all the world’s DOI activity. Their records (about 180 million of them as of March 2026) include metadata about individual research works (articles, datasets, books, grants, preprints, peer-reviews), as well as journals, funders, and even their own member organizations.

That means if you’ve ever clicked on a DOI to read an article, look up a grant, or cite a dataset, it’s pretty likely Crossref has a metadata record for whatever it was you were interested in. Whether you want to know more about a single article or you’re interested in conducting metaresearch at scale, Crossref’s records can help you get a start on answering practically any question you can come up with.

While it’s true that Crossref provides free, open access to these records as a service to the scholarly community, some of the more useful ways of accessing that data have pretty technical-sounding names like API and can feel arcane and unapproachable. In this series of articles, I hope to dispell that apprehension by focusing on the bare necessities needed to get started doing meta-research. Along the way, you’ll learn how Crossref makes their metadata records available to the public and how to write simple-to-intermediate programs that engage responsibly with that service to be able to answer the following types of questions:

• What grants have recently been awarded in my field of study?
• Who funded this clinical trial I am reading?
  
• How many retractions has a journal/publisher issued in the past year?
• What’s the copyright license for this book chapter? Or, what’s the most common copyright license this journal/publisher issues?
• Are publishers publishing fewer books on my topic of interest now than in the past?
• Have the average number of references provided by authors in my area of interest changed over time?

Next article: What’s an API?

Scholarly Metadata

Metadata can be thought of as data about some other piece of data. Or in other words, metadata is information that describes the data it’s attached to.

In the case of a scholarly work, such as a journal article, the ‘data’ is the article, and the metadata includes all the little pieces of information you may need to identify it. For example. if you wanted to cite that article, you’d need:

• Title
• Authors
• Publisher
• Date of Publication

…etc—all metadata. But metadata can include much more—for example, the digital object identifier (DOI), or the exact time that that identifier was registered online, or the copyright license the publisher has assigned to it.

A particularly important piece of metadata in the digital age is “where is the digital location of this data?” Or, similarly, “What website do I need to go to so that I can view or download this data?”

Other pieces of metadata may include Relationships:

• Who funded this work?
• Which paper(s) cited this book?
• Does this medical diagnostic guidance document update a prior guidance document?
• Which hospital was this author affiliated with when she wrote this article?

Digital Object Identifier (DOI)

DOIs are an internationally standardized system of identifying ‘digital objects’, which in practice are mostly research works such as journal articles, in such a way that the identifier never has to change, even if the object moves location or transfers ownership. The international standard is facilitated by the DOI foundation which governs multiple Registry Agencies who work with publishers to assign DOIs. Crossref is one of these agencies.

Because there are many agencies, and agencies may come and go, DOIs are minted with the domain of the DOI foundation, and look like this: https://doi.org/10.1162/qss_a_00022 and always start with https://doi.org/ (older DOIs may also start with https://dx.doi.org/), even though this DOI was registered by Crossref, whose domain is https://crossref.org

Here’s some more information on the background of DOIs and why they are needed (from the DOI Foundation)

You can see the full list of registry agencies including those other than Crossref here.