For some publications, a database or SGML/XML tagged data provides the best management scenario for the data to be published.

A database provides structure that is identified by the real-world relationships between data residing in tables of records and fields in the records. Content is determined by the field and record that data is stored in while context is provided by the relationships defined between the tables. This allows development of processes indirectly related to the actual publication of the data.

For instance, reports of the data ordered in ways not found in the publication, such as phone book entries ordered by telephone numbers. Another example is questionnaires and labels generated to mail to listees in a directory where the generated data needs to be sorted by state and zip code.

Typical publications that work well with a database approach are price catalogs, directories, thesauri, and abstract journals with indexes. These publications are often “living” collections of data that cumulate or get revised on frequent basis.

A mark-up tag approach (more often than not XML now) stores data organized in publication order in file(s) with the tags denoting the relationships but only as they relate to one view of the data—the publication view. Content is determined by the naming of the tags applied to the data while context is provided by the hierarchical structuring of the data.

Other uses of the data required that it be extracted and processed.

Typical applications utilizing the mark-up tag approach are publications that are more text-oriented, such as text books, technical documentation, monographs, and treatises. They are often one-time or infrequently revised publications but whose overall structure might be used in series of publications on different subjects.

It should be noted that, though these two approaches seem quite different, they are in fact logically equivalent to each other. Both store data in a way which provides both context (relationships) and content.