Data Conversion - Information Systems Reengineering, Integration and Normalization - page 146

However, this approach is proprietary, language oriented (not user friendly), and too
expensive to adopt.

4.4 Logical Level Translation Approach

This approach is similar to the interpretive approach but proposes the reduction of
storage and physical costs and without the need for specialized description languag-
es. Instead, it considers only the logical level of data representation. For example,
Shoshani ( 1975 ) used a source definition of the network database and the network
DML to read the data from the network database and store it in a convenient, in-
termediate target. The intermediate target format was then read and stored in the
relational database using the definition of the relational database and the relational
DML as illustrated in Fig. 4.6 .
There are two parts to this problem: unloading the data from the nonrelational or
relational database, and uploading the data into the relational database or from the
relational to object-oriented or XML database. The two steps are independent, since
most vendor load utilities accept a simple flat file as input. Any available utility that
can read the source database and creates a flat output file can be used for this pur-
pose. These output sequential files should be reorganized into a logical sequence for
the uploading process after the generation of the new database definition. Generally
the load utility can be applied in the upload process.
The logical level approach is more commonly used in the industry because it is
easier to implement than the others. The later sections describe using the logical
approach to convert data from a network database to a relational database, from a

Definition
of source

Definition
of mapping

Definition
of target

DBMS query
function

DBMS insert
function

Reformat

Intermediate
target

Source Intermediate Target