How to follow research paper updates with https://api.crossref.org/works API?

DOI metadata of research papers are changing changing. If you want to keep up with metadata changes, you can use the https://api.crossref.org/works API to implement an automated process.

1. Create 2 tracking variables and initialize them with very low values:

upd_date = 1950-01-01
upd_page = -1 

Step 1 is one-time only. If your automated process gets interrupted, you should restart it from Step 2.

2. Fetch DOI metadata updated of the next day and sorted by update time with next cursor enabled:

date = upd_date + 1 day
page = 0
https://api.crossref.org/works?filter=from-update-date%3A{date}%2Cuntil-update-date%3A{date}&sort=updated&order=asc&cursor=*

3.1. If (page > upd_page):

Process all entries in the response.
upd_page = page
page = page + 1

3.2. Else (page <= upd_page):

page = page + 1

4.1. If next-cursor exists in the output:

https://api.crossref.org/works?filter=from-update-date%3A{date}%2Cuntil-update-date%3A{date}&sort=updated&order=asc&cursor={next-cursor}
Go to step 3

4.2. Else (next-cursor does not exist in the output):

upd_date = upd_date + 1 day
upd_page = -1 

Parameters and filters used in the API:

from-update-date:{date} - Limit deposited.date-time > date in the output
until-update-date:{date} - Limit deposited.date-time < date+1 in the output
sort=updated - Sort output by deposited.date-time
order=asc - Sort deposited.date-time in ascending order

⇒ DOI API for Python - Crossref Commons

⇐ crossref.org API - /works/{doi}/transform/{mime}

⇑ Crossref.org DOI Database

⇑⇑ DOI - Frequently Asked Questions