27-10-2012, 01:48 PM
Metadata for Electronic Information Resources
ABSTRACT
The rationale for cataloging and indexing of electronic information is much the same as for print materials. Cataloging
and indexing provide a surrogate for the item, which facilitates resource discovery and access. But, what has changed in the
electronic information environment is the terminology. In the Internet environment, the terms cataloging and indexing have
been replaced with the term metadata. Metadata is often defined as data about data or information about information. The term,
which originated with the data and computer science communities, is now in general use for the cataloging and indexing of
electronic information sources. Metadata serves three general purposes. It supports resource discovery and locates the actual
digital resource by inclusion of a digital identifier. As the number of electronic resources grows, metadata is used to create
aggregate sites, bringing similar resources together and distinguishing dissimilar resources. There are a variety of metadata
schemes that serve different purposes for different object types, subjects and audiences, including the Dublin Core, Metadata
Object Description Schema (MODS), the Global Information Locator Service, the Text Encoding Initiative Header, the
Encoded Archival Description, the Content Standard for Digital Geospatial Metadata, the Data Documentation Initiative, and
the draft Technical Standard for Still Images. A metadata scheme has three components semantics, content and syntax. An
extension adds elements to an existing scheme to describe a particular resource type, handle material on a particular subject,
or address the needs of a particular user community. Profiles are subsets of a larger scheme that are implemented by a
particular user community. Metadata can be embedded in an electronic resource or stored in a separate file. A growing number
of tools, both open source and commercial, are available to create and edit metadata. Creation may be done manually or by
metadata generators that extract key information from the object. Metadata harvesters capture metadata records that have
already been created using the ‘shared cataloging’ model. While many projects aimed at having metadata created by the object
s author, this has proved to be difficult to implement. An alternative is to have a core set of metadata created by the author
with editing and quality control performed by a librarian or editor who has a view of the whole collection. With disparate
metadata schemes, ensuring that information collected in a specific scheme by one organization for a particular purpose can
be exchanged, transferred or used by another organization for a different purpose becomes an issue. Metadata frameworks,
crosswalks, and registries are ways to achieve interoperability. Use of controlled and uncontrolled vocabulary terms is
encouraged, particularly within specific subject domains. However, most metadata schemes do not dictate the use of a
particular controlled vocabulary but instead allow the vocabulary scheme to be defined within the syntax.