We’ve recently started collecting events from around 10,000 more domains.
Most research mentioned online uses the publisher URL rather than the DOI, however we would still like to capture these as events in Event Data. We therefore keep a mapping between DOI prefixes and landing domains. When we recognise one of the domains in one of our sources (e.g. on Wikipedia, Twitter, or a blog post) we try to match the URL back to a DOI. The allows us to capture many more events than if we simply looked for a DOI.
Over time, prefixes are created and transferred between our members, and website domains change or are added. We have recently been through the process of updating prefix-domain mapping, which hasn’t been done since 2018. We found around 10,000 new domains, half of which are for prefixes not previously included. To see which domains and prefixes are included, see the domain-decision-structure artifact which can be found by following the links in our user guide.