HP Integrity NonStop J-Series - Manual (Page 234)

Sample Project-Tracking System
Identifying Data Items
After they have been identified, you can ask specific questions
about each data item; for example:
• Are any of the data items associated with more than one class
of data? If so, is there a reason for this redundancy?
• Are any data items repeated within the same class of data? If
so, can the data base be normalized?
Examination of the data items associated with "projects" and
those associated with "employs" shows that both classes contain
an employee number, an employee name, and a department number.
Note that in the case of the "projects" data, the employee
information is that of the project manager.
Redundant data such as this can use extra storage and complicate
updates. If, for example, a project manager's department number
changes, you must correct both the "projects" and "employs" data,
leading to extra work and possible mistakes.
A simple solution is to remove the employee information, except
for the project manager's identification number, from the
"projects" data. You should leave the project manager's employee
number with the projects data to provide a link to the "employs"
data. With this link, detailed information about the project can
be maintained separately from that of the employees, yet their
association can be preserved.
Notice also that the same data item (an employee name) appears
both in the "events" data and the "employs" data. In this case,
the employee information is not only redundant; it is also
confusing. If more than one employee has the same name, you
cannot clearly identify the employees assigned to an event. To
eliminate the confusion, replace the employee name in the
"events" data with an employee number. Although this solution
eliminates the confusing data, it does not remove repeating
information from the "events" data. Since more than one employee
can be assigned to an event, more than one "emp-no" could be
associated with each event. The next step in the development
process provides a solution to the problem of repeating data.
12-6