Introduction

Descriptive metadata forms the core of any library, archive or repository, analogue or digital, where it plays the indispensable role of ensuring that the contents of its collections can be discovered and unlocked. It is the type that is most visible to end-users, for whom it provides signposts through the physical or virtual shelves, making it possible for them to locate what they need and to avoid what they do not. Without adequate descriptive metadata a digital library is almost unusable, the time and effort expended in creating and maintaining its collections going to waste.

Its central position in a library's work necessitates a place of commensurate importance in its overall metadata strategy. For this reason alone it should certainly adopt the principles outlined in Chapter 2. In particular, it should adhere to Principle 7 (‘Use standards whenever possible’) and so follow established practices which have developed within professional communities and been proven robust and viable. Descriptive metadata should also take particular note of Principle 4 (‘Ensure interoperability’) so that it can readily be shared with others and so widen its potential audience well beyond the boundaries of the collections within which it is created.

This chapter concentrates on the semantics of descriptive metadata, the containers that hold it and how they are labelled and described. It introduces three metadata schemes, DC, MARCXML and MODS, all of which are compatible with METS and all of which have established themselves firmly within the digital library community. These are generic schemes which can accommodate descriptive metadata for a wide range of digital objects; more specialised ones which can either replace or supplement them for specific collections will be discussed briefly at the end of the chapter.

Dublin Core

Dublin Core, or DC as it is often called, owes its name to the city of Dublin, Ohio, where in 1995 a group of experts from computer science, libraries, archives, museums and online information services met to put together a standard for describing resources on the internet. They were attempting to emulate the way in which the MARC cataloguing standard had opened up new avenues for discovery in libraries by enabling the sharing and exchange of metadata in a logical machine-readable format.