Creating globally unique and persistent identifiers (PIDs) and publishing accompanying metadata to facilitate open science
Globally unique, persistent and resolvable identifiers (GUPRIs) are key to enable and propel open science forward globally. However, the scientific community is currently relying on centralized registries like ORCID (for people) and DOI (for publications) which are quite hard to add to for an individual scientist. This is not so much a problem for these examples, but if you want to enable more specific identifiers for digital assets, e.g. for medical studies or scientific datasets, it become really hard to ensure that identifiers are GUPRIs. Decentralized protocols like Ethereum and IPFS offer a unique chance to address this, however pure IPFS (e.g. encoding metadata as JSON-LD or even better CBOR-LD to obtain a CID) would suffer from not being able to update or version the metadata.
How it's made
We are looking to build a prototype, probably using Ceramic but also open to other options, of a publishing solution that allows us to mint GUPRI identifiers in a decentralized way. As test dataset we will be using an actual published medical study on COVID as example that has been used by the European Medicines Agency for guidance on clinical treatment. We will be using JSON-LD (and schema.org) to encode key metadata and use that to mint the initial digital object identifier and publish metadata around it. There are many requirements, we will not be able to satisfy them all in a weekend but we can make a choice at the start which one we focus on.Technologies used