Databases Reference
In-Depth Information
The objective of logical design is to transform the conceptual
schema by adapting it to the data model that implements the DBMS
to be used (usually relational). In this stage, a logical schema and the
most important users views are obtained.
·
The objective of physical design is to achieve the most efficient
implementation of the logical schema in the physical devices of the
computer.
·
During the last few years, there have been many attempts to offer a more sys-
tematic approach to solving design problems. In the mid-1980s, one of those
attempts was design automatization through the use of computer-aided soft-
ware/system engineering (CASE) tools (see Chapter 13). CASE tools con-
tributed to spreading the applications of conceptual modeling and
relaunching DB design methodologies. While it is true that some CASE
tools adopted more advanced approaches, many continued to be simple
drawing tools. At times, they do not even have a methodological support or
are not strict enough in their application. As a result, designers cannot find
the correct path to do their job [17]. Furthermore, the models the tools gen-
erally support are logical models that usually include too many physical
aspects, in spite of the fact that the graphic notation used is a subset of the
E/R model.
New (object-oriented) analysis and design techniques, which at first
focused on programming language and recently on DBs [18, 19], have
appeared in the last decade. Those methodologiesBooch method, object-
oriented software engineering (OOSE), object modeling technique (OMT),
unified method, fusion method, Shlaer-Mellor method, and Coad-Yourdon
method, to name some important examplesare mainly distinguished by
the life cycle phase in which they are more focused and the approach adopted
in each phase (object-oriented or functional) [20]. A common characteristic
is that they generally are event driven.
The IDEA methodology [21], as a recent methodological approach, is
an innovative object-oriented methodology driven by DB technology. It
takes a data-centered approach, in which the data design is performed first,
followed by the application design.
1.3
The New DB Generation
Many nontraditional applications still do not use DB technology because
of the special requirements for such a category of applications. The current
