Database and Expert System Technology - Information Systems Reengineering, Integration and Normalization - page 62

2.5.1 Data Type Definition Graph

XML started in 1998 as a new data standard on the Internet. XML documents can
be stored in a native XML database or an XML enabled database. The former is an
XML oriented database management system. The latter is relational database with
an XML Application Program Interface (API).
To design an XML database, one needs to construct an XML Schema Definition
(XSD) Language Graph in the form of a hierarchical containment, starting with a
root element on top of other elements. An XML schema can be stored in a Data Type
Definition (DTD) or an XSD.
Given the DTD information of the XML to be stored, we can create a structure
called the Data Type Definition Graph (Funderburks et al. 2002 ) that mirrors the
structure of the DTD. Each node in the Data Type Definition Graph represents an
XML element in rectangle, an XML attribute in semicycle, and an operator in cycle.
They are put together in a hierarchical containment under a root element node, with
element nodes under a parent element node.
Facilities are available to link elements together with an ID and Identifier Refer-
ence (IDREF). An element with IDREF refers to an element with ID. Each ID must
have a unique address. Nodes can refer to each other by using ID and IDREF such
that nodes with IDREF referring to nodes with ID.
An XML document is in a hierarchical tree structure. Every XML document
must have one root element. The root element is in the highest hierarchical level.
The root element contains all the other elements and attributes inside of it. Other
elements are in hierarchical order such that they are in relative parent or child node.
That is, the relative higher level is the parent node and the relative lower level is
the child node.
An element is the basic building block of an XML document. An element name
must start with a letter or underscore character. An element can have subelement un-
der it. An empty element does not have a subelement. Between element and subele-
ment, there are declarations that control the occurrences of subelements. For exam-
ple, one can define element instances in a Document Type Definition (DTD) with
an Occurrence indicator. For example, the “*” operator identifies “set” subelements
that can occur from zero to many times under a parent element. The “+” occurrence
indicator specifies one to many times occurrence under a parent element. The “?”
occurrence indicator specifies zero to one time occurrence under a parent element.
Attributes give more information about an element and reside inside of the ele-
ment. Attributes can further define the behavior of an element and allow it to have
extended links through giving it an identifier.
For example, the following is a Data Type Definition Graph with root element
Patient_Record. In Fig. 2.11 , the Data Type Definition Graph has a root element Pa-
tient Record. Under the root element Patient Record, there is an element of Patient.
Element Patient has a sub-element Record folder. The Element Record folder has
one sub-element, Medical record. Element Medical record has the sub-element AE
record, sub-element Ward record, or sub-element Outpatient record.