What is metadata?
Metadata is typically defined as "data about data," which can be encoded within the object itself or separately.
Metadata in the context of the digital library can be seen as parallel to what libraries have done for years in creating the card catalogue, and later the online public access catalogue (OPAC) with records encoded in MARC. An important difference in this analogy, however, is that metadata in the digital library context is typically much closer to its resource than the library catalogue record is to a physical book. In the digital library, an object's metadata is often embedded in or otherwise tightly integrated with the resource itself, so that in a sense this kind of metadata can be seen as part of the resource itself, as an enhancement of the original resource. Digitization makes this close association between metadata and resource possible in ways not typically possible with analogue materials, and so the addition of robust metadata to digital assets is an essential means to leveraging the digital medium in the execution of a digital project.
Unlike their physical counterparts, digital resources are not stored in a static container; instead, they are dependent on mechanical devices which run software. This poses several challenges for the creators and distributors of digital information. The first challenge is whether the software used in a digital project will be supported in the future. To relate that to the realm of human language, imagine publishing an advertisement today in either Latin or Sanskrit. Most people would not be able to read, understand, or process the information. As a means of communication it would be quite unsuccessful. A further concern could be whether the hardware will still function properly in the future. Libraries have witnessed this failing as information was stored on microfiche and over the years the microfiche readers fell into disrepair. So how are these problems overcome?
Digital libraries have developed resources that are independent of hardware or software by using descriptive terms and rules known as metadata schemas. Metadata allows one to record information that organizes and describes digital resources. There are many metadata schemas in use, some of which are designed for specific formats. On this site, Digital Library Services and the Metadata Subcommittee will suggest schemas according to the type of digital project to be undertaken.
Types of metadata
Metadata can be broken down into three groupings. These are descriptive, structural, and administrative.
Descriptive Metadata: This form of metadata describes the intellectual and physical properties or content of a digital object. Examples of descriptive metadata include: titles, authors, and editorial notations.
Structural Metadata: This form of metadata describes the component parts of a digital object in relationship to the whole digital object. Examples of structural metadata include: file types and versions, as well as chapters or indices.
Administrative Metadata: This form of metadata is about the record itself; this includes its creation, modification, location and relation to other records. Administrative metadata also contains technical information regarding software or hardware used as well as information about preservation.
How is metadata applied?
Because interoperability is a key function of metadata, the creation of metadata standards for different digital media has been a key task in recent years, and several new standards have been adopted recently. A list of some of the more important metadata standards is found on the Metadata Standards page. Interoperability is further possible through use of Extensible Mark-up Language or XML, which allows the information in a metadata standard to be more easily shared and managed. While there are a large number of metadata standards in use, Digital Library Services and the Metadata Subcommittee encourage use of the following standards on the Metadata Resources page based on the original format of the collection to be digitized. In some cases, multiple metadata standards will be needed to describe a digital project. While standards provide a common vocabulary for the description of resources, any given context will have unique requirements, so it is typically necessary for institutions to combine basic elements from different standards to create a local scheme for a digital library. One such scheme is currently in development at Washington University.
Metadata at Washington University
Washington University's Digital Library encourges digital projects include a minimal set of metadata elements. We encourage you to consider other elements for current and future description and retrieval.