HP 3000 Manuals HP 3000 Manuals
Data Dictionaries
Just as an ordinary dictionary is a collection of definitions of words, a
data dictionary is a collection of definitions and descriptions of data
that resides on a computer system. In a dictionary, the smallest unit of
information is a word, while in a data dictionary the smallest unit of
information is a data element.
You can compare a data dictionary to the file card system in a public
library. Each card in the file contains a description of a book in the
library, and lists the title, author, publisher, location in the library,
etc. of the book. The card itself is not the book, but only represents
and describes it.
Like the file cards, a data dictionary does not contain the data itself,
but contains metadata--data about data. This metadata can be
descriptions and definitions of various kinds. It can describe such
things as:
Data: Names and definitions of data
elements
Data Relationships: How data is related to other data
Data Responsibility: Who is responsible for what data
Organizational Structure: The information flow, who uses the
data
Location Information: Where files, programs, and reports
reside
Security Information: Who has access to what data
A typical example of a piece of metadata is a data element called "SSN",
which represents a piece of data--a social security number. The social
security number itself does not reside in the dictionary, but a
description of that piece of data does. For instance, the data
dictionary might tell you the name of the data element, the storage
length, the display or output length, the type of data (numeric or
character), its sign if the element is numeric, which database or program
that data resides in, and possibly which departments in the organization
use and maintain that data.
The data dictionary, therefore, serves many purposes. You can use it as
a quick directory to the information that resides on a computer
system--where to go to get pieces of data. You can also, however, use it
as one of the primary means for ensuring consistency of data definitions
and preventing data redundancy. This means that programmers and
developers may be required to check the data dictionary for data elements
that already exist on their system before they design a new program.
Therefore, if a data element already exists on the system describing a
social security number (for example, "SSN"), the dictionary reports this
information and does not allow the programmer to add a new element with
the same name. This helps an organization to save time in program
development by using data definitions that already exist. It also saves
data storage space by preventing data redundancy and helps to standardize
data definitions within an organization.
For a more detailed introduction to what data dictionaries are and how an
organization can use a data dictionary, see the HP primer entitled
Managing Your Information Network: A Data Dictionary Primer (5958-8527).