Hello Crossref community,
I am developping a tool for TDM for retrieving full text. I have read through the documentation of the Crossref API and the different other endpoints you offer but I am having trouble retrieving full text from publishers with a paid subscription. If I pay a subscription to access paid full text content on Elsevier for example, how do I use the Crossref API to access this full text? The metadata returned by this API contains full text links, but the content is inaccessible from simply accessing the URLs. They usually require credentials, such as an API key in a specific format in the header of the request.
What is the recommended method for retrieving full-text from paid for publishers? Do you know of any examples of a request for paid-for content?
Thank you, please let me know if I have provided enough details in my question,
That’s a question you’d need to direct to the respective publishers.
Crossref only indexes metadata, not full text, and we don’t have anything to do with access control or authentication. Each publisher handles that differently.
Publishers may opt to supply full text URLs in the metadata records they submit to Crossref, and these will be tagged with an “intended-application” which is typically either “similarity-checking” or “tdm” (text and data mining).
The “similarity-checking” links are only used by Turnitin, as part of our Similarity Check service.
The “tdm” links are intended for use by anyone who has a text and data mining agreement with that particular publisher, but you’d need to establish access directly with the publisher. Or, in cases where the full text content is covered by an open access license which applies text and data mining, no authentication is typically necessary.
Please let me know if you have any other questions.
Thank you for the quick and thorough answer!
If I understand correctly, the full-text links provided by the publishers to the CrossRef metadata is not guaranteed to lead directly to a file download of a full-text?
They should lead to the full text only if you’ve already established tdm access through that publisher.
Unless the full text content is already open access or otherwise freely available, you should expect to need some kind of authentication, whether that’s an API key, IP auth, user credentials, or similar.
Thank you I understand a lot better now.
I just want to make sure I got all that, let me present an exemple of the workflow of downloading a full-text with the tool:
- Fetch the CrossRef metadata.
- Check if the publisher’s TDM license is white listed.
- Use the necessary authentification depending on the publisher.
- Download the article from full-text link.