Relational Database

A relational database stores all information in tables, and a
user can manage data by direct manipulation of these tables
without reference to other constructs.  Any piece of information
in a relational database can be accessed directly by referring to
the table number, key value, and column name.  The structure of a
DBMS also permits the manipulation and merging of data fields
according to predefined operations.  Basic relational operators
(select, project, and join) each produce a new table by combining
one or more tables on the basis of shared data fields.  These
operations include "joins", "nested-loops joins", and "sort-merge
joins".  "Joins" are simply a way of merging two tables.  A
"nested-loops join" compares every row in one table to every row
in another, and combines the qualifying rows.  A "sort-merge
join" sorts each table on the columns that will be compared, then
scans the tables and joins the qualifying rows (Robie, 1988). 
While operations such as these can provide a skillful user with
tools that permit the creative comparison and combination of data
fields, they have limited capabilities in accessing and
manipulating idea entities. As a result, a DBMS structure does
not represent the best strategy for an integrated information
system that contains widely fragmented and diverse idea
resources.

An electronic preservation information application based on a
DBMS structure has been developed by the National Park Service,
and this system was evaluated as part of this project.  This
system uses an off-the-shelf software program called dBASE III
Plus (Ashton-Tate, Inc.).  The application is described in
Section 3.1.0.  

<cfile20>   Database Management Systems (DBMS)
<dfile13>   HSPD