Common Logical Data Model: Basis for Global ITS Innovation
among associations, just as we have shown
relationships among classes. While this may
seem unnecessary for this example, there
are other cases where this can become quite
useful. For example, by defining association
classes, one could create a formal ontology
that defines that one person can be the
“wife” of another person and, if so, then the
other person is the “husband” of the first.
Further, both the “husband” and “wife”
associations can be formally defined to be
subclasses of the “spouse” property. A
computer system equipped with this
definition would then recognize that any
husband can also be called a spouse.
of tables that have to be managed. The
logical data model escapes these types of
constraints and should be designed to reflect
real world artifacts as closely as possible.
The logical data model reflects the
conceptual data model as closely as possible.
However, whereas the conceptual data
model will define equivalent terms and
other artifacts that are important to capture
for human discussions, the logical data
model limits or omits this type of
redundancy. In addition, the logical data
model defines additional detail regarding
the data including the units in which
measurements are made and the level of
privacy that should be associated with the
data.
The conceptual data model therefore
becomes a central resource for deep
learning in being able to interpret written
text. But it falls short in providing a concise
view of the information for defining future
interfaces or for translating among existing
physical data models. For that, we turn to
the logical data model.
Standardization Process
A CCESS TO THE M ODEL
One of the challenges of producing a model
that is intended to represent and be used by
the entire industry is that it must be readily
available to receive inputs from a massive
stakeholder community. This necessarily
requires that the community:
L OGICAL D ATA M ODEL
The goal of the logical data model is to
provide a “Rosetta Stone” for the industry. It
allows data implemented according to one
data format (i.e., physical data model) to be
transformed to any other data format by
formalizing the transformation to a common
data format.
•
•
Typically, SDOs charge fees for one or both
of these functions. For example, ISO offers
free participation in the development of
standards but charges for access to the
resulting
standards.
OMG
charges
membership fees for contributing to the
standards development but offers the end
product to the community for free. The IETF
allows for free contributions and free access
By their nature, physical data models deal
with constraints related to the environment
that they are intended for. For example,
physical data models for communication
protocols often attempt to compress data to
minimize the size of the data that has to be
transmitted. Physical data models for
databases often try to minimize the number
IIC Journal of Innovation
Is able to readily access the model
Is able to readily provide input to the
model
- 40 -