A Decentralised Persistent Identification Layer for DCAT Datasets

The Data Catalogue Vocabulary (DCAT) standard is a popular RDF vocabulary for publishing metadata about data catalogs and a valuable foundation for creating Knowledge Graphs. It has widespread application in the (Linked) Open Data and scientifc communities. However, DCAT does not specify a robust mechanism to create and maintain persistent identifers for the datasets. It relies on Internationalized Resource Identifers (IRIs), that are not necessarily unique, resolvable and persistent. This impedes fndability, citation abilities, and traceability of derived and aggregated data artifacts. As a remedy, we propose a decentralized identifer registry where persistent identifers are managed by a set of collaborative distributed nodes. Every node gives full access to all identifers, since an unambiguous state is shared across all nodes. This facilitates a common view on the identifers without the need for a (virtually) centralized directory. To support this architecture, we propose a data model and network methodology based on a distributed ledger and the W3C recommendation for Decentralized Identifers (DID). We implemented our approach as a working prototype on a fve-peer test network based on Hyperledger Fabric.