Data Model – Notes from Think Big

In last weeks Think Big DW Conference 2006, Stephen Brobst the CTO at Teradata gave a talk on The Sins of Denormalization.  It was a great little chat with the audience.  In the talk he discussed the Data Model.  Before I start, I thought I would jump on Wikipedia to get a full definition first, so here it is.

A data model is a model that describes in an abstract way how data is represented in a business organization, an information system or a database management system.

This term is ambiguously defined to mean:

  1. how data generally is organized, e.g. as described in Database management system. This is sometimes also called "database model"
  2. or how data of a specific business function is organized logically (e.g. the data model of some business)

So I thought I would take the introductory ideas that Stephen brought and add some detail from some research and some other wrappers.

Logical Data Model

 

In computer science, a logical data model is an abstract representation of a set of data entities and their relationships.

 

A logical data model typically includes the primary and foreign key
attributes of the modeled entities but omits non-key attributes.
Depending on the notation used and the level of abstraction, a logical
data model may use abstractions of actual types and may represent
physical tables as relationships or as aggregates within a larger
entity.

 

The purpose of a logical data model is to facilitate analysis of the
function of the data design. It is not intended to be a full
representation of the physical database. It is typically produced early
in system design, and it is frequently a precursor to the physical data model that documents the actual implementation of the database.
– Wikepedia definition.

 

Physical Data Model

A physical data model (a.k.a. database design)
is a representation of a data design which takes into account the
facilities and constraints of a given database management system. In
the lifecycle of a project it is typically derived from a logical data model,
though it may be reverse-engineered from a given database
implementation. A complete physical data model will include all the
database artifacts required to create relationships between tables or
achieve performance goals, such as indexes, constraint definitions,
linking tables, partitioned tables or clusters. The physical data model
can usually be used to calculate storage estimates and may include
specific storage allocation details for a given database system.

– Wikepedia definition.

Semantic Data Model
Conventional data models are not satisfactory for modelling data base
application systems. The features that they provide are too low level
and representational to allow the semantics of a data base to be
directly expressed in the schema. The semantic data model (SDM) has
been designed as a natural application modelling mechanism that can
capture and express the structure of an application environment. The
features of the SDM correspond to the principal intensional structures
naturally occurring in contemporary data base applications.
Furthermore, facilities for expressing derived (redundant) information
are an essential part of the SDM; derived information is as prominent
in an SDM schema as is primitive data. The SDM is designed to enhance
the effectiveness and usability of computerized data bases. It can
serve as a formal specification and documentation mechanism for a data
base, can support a variety of powerful user interface facilities, and
can be used as a tool in the data base design process.  A good example
of the physical equivalent is the Business Objects Universe.

Semantic translation is the process of using semantic information to aid in the translation of data in one representation or data model
to another representation or data model. Semantic translation takes
advantage of semantics that associate meaning with individual data elements in one dictionary to create an equivalent meaning is a second system.

  – Wikepedia definition.

I really liked the inclusion of the Semantic model.  This is a great tool when working with the a variety of Business Intelligence tools and easing the integration and performance.  What are your experiences.

 

Leave a Reply

captcha *