In the databases, a document is basically correlated with having the same structural data for common segments. Each of the fields can represent itself multiple times in a document with the different data types.
