XML and MARC

In the field of information and library science (LIS), one of the major challenges of the last twenty or so years of the twentieth century was converting the data about items held in a library from index cards--on which librarians had simply typed the metadata (author, title, publisher etc...)--to computers. A few different solutions have been developed for this problem have been developed, including the Dublin Core Metadata Initiative (DCMI), the Web Ontology Language (OWL), and MAchine Readable Cataloging (MARC).

The main way that MARC works is that it assigns all fields, such as author, title etc... as an element, assigns tags and indicators as attributes, and it treats subfields as subelements. For example, an item might be assigned a datafield with a number, while each piece of metadata about the item has been assigned an element, such as a, b, c, etc... When representing an item, for example The Grapes of Wrath by John Steinbeck, the system would assign the book an arbitrary number and have a preassigned definition of title, author, date, publisher, and ISBN as the elements. The record might look like this:

< datafield tag="670" ind1=" " ind2=" ">
< subfield code="a" > The Grapes of Wrath, c1939: < / subfield >
< subfield code="b" > (John Steinbeck) < / subfield >

MARC is especially beneficial because as an XML file it can share data with other DTDs, such as the DCMI. MARC's official website explains in further detail of the various ways to utilize the DTD, as well as examples of how it marks up various references.