The web provides arguably the most broadly accessible, and provenly scalable infrastructure for resolving identifiers on the planet. The underlying DNS infrastructure provides a scalable means of decentralized allocation of identifiers at the domain level. More than a decade of experience is available on standards and techniques for using http URIs to identify concepts and real world entities, not simply web resources.
Consistent use of standards can be applied to both ‘unstructured’ information i.e. documents, and even moreso, to structured data. By applying the right metadata, and semantics, we can make joins to discover, describe, and share both documents and data.