What is Logical Data Modeling?

by Jun 25, 2021

Logical data modeling is a vital procedure in database design and implementation. It is the second of three stages of data modeling, forming the bridge between conceptual and physical data modeling. During the process, data modelers construct logical data models.

On this page:

What Is a Logical Data Model?

A logical data model is a graphical representation of an organization's information requirements. One of three types of data model, logical data models are used to define data's structure and display the relationship between different entities. 

They provide more context than what is captured within conceptual models – adding attributes, assigning data types, and specifying requirements for the information. Such requirements include primary keys and business rules that may apply to the information.

As with conceptual data models, logical data models still utilize natural business language.

Who Constructs Logical Data Models and Why?

Typically created by Data Architects and Business Analysts, the logical model is less abstract than a conceptual model and helps teams understand the data’s details and the logistics of building and implementing a new system. As with conceptual models, they are platform independent and can therefore be created without committing to an implementation path.

This provides organizations with the opportunity to evaluate their current state, and decide whether they need to invest in new technologies to target the desired future state. It may also become apparent that the “ideal” future state is out of scope, and that objectives need to be revised. 

Logical Data Model

What Is the Goal of Logical Data Modeling?

The end-goal of a logical data modeling is to construct platform-independent data models to represent business entities and their relationships in third normal form (3NF). The resulting logical data model should provide a business user-friendly view of data structures within a data asset – i.e. a database.

As the link between conceptual and physical data models, the logical model should be detailed enough to kick start the physical data modeling process.

As well as contributing to data asset design, organizations can use logical data models to inform other enterprise data initiatives such as data governance. For example, organizations can harvest business terms from logical data models to populate their business glossary. 

Other Types of Data Model

Here’s a quick look at the other types of data models:

www.youtube.com/watch

Conceptual data models can be defined as high-level models that provide an in-depth view of business concepts. They are useful for identifying key business and system entities and establishing relationships between them. 

Conceptual models are platform-independent, delaying the choice of database platform until later in the development process. These models are usually created for a business audience.

Physical data models describe how the information and details identified through the conceptual and logical modeling phases will be implemented using a particular database platform. The physical models define the database schema as well as the design elements such as tables and keys needed to develop a system.

Once the physical model is completed, the database team should have all the information they need to implement the design on the given platform.

Can I Skip Logical Data Modeling?

In short – no. Constructing logical data models is a fundamental best practice for data asset design. 

Some teams may attempt to bypass the creation of a logical data model and move straight from a conceptual model to a physical model and implementation. They may try this due to time or resource constraints. Perhaps they feel it is an unnecessary step and are prepared to develop a database based purely on a conceptual model.

This approach is not recommended and can lead to more work in the long run, obliterating any time savings that were accrued by ignoring the logical model. 

Without a logical model that fleshes out the details of the prospective database, the business objectives identified and codified in a conceptual model risk being lost or misinterpreted.

It can be extremely challenging to accurately turn business requirements directly into the constructs required by a specific database platform. This difficulty can result in physical models that do not reflect the needs of the business. 

Another reason not to skip the logical modeling phase is that during model creation, it may become apparent that the business concepts attempting to be addressed are out of scope regarding current systems and resources. A logical model can point out systems and processes that need to be modified or improved.

When Should I Build a Logical Data Model?

Ideally, data models should be built linearly – starting with a conceptual model and progressing it through the logical and physical stages. This means logical data models should be built after the conceptual data modeling phase. 

However, this is not always the case. Often, data professionals may reverse engineer databases and scripts to create logical data models. This can help organizations better understand an existing data asset that they may not have a current model for.

This is particularly useful when an organization wishes to demonstrate the databases structure, redesign an existing data asset, or even create product documentation.

How Do I Build a Logical Data Model?

In the standard data modeling workflow, the creation of a conceptual data model is a prerequisite for building a logical model. During the construction of the logical model, details are added to fully document how the ideas outlined in the conceptual model will be represented in a database. 

While the conceptual model is concerned with abstract business requirements, the logical model begins to fill in the detail. It defines the structure of data elements and the relationships between them. Some of the types of information that are defined in logical models are primary, foreign, and alternate keys. Attributes like the format and length of data elements are defined in a logical model. 

The way the data elements will be represented in a database is left to the physical model and are not part of logical modeling. Tables, columns, and other physical objects used to implement the original business concepts are not addressed in this phase of data modeling.

Using a dedicated data modeling tool can give teams a big advantage over attempting to complete the process manually. Simple models may be developed manually, but as the complexity ramps up, it becomes increasingly difficult to capture all relevant information and leaves teams open to making mistakes.

Subscribe to IDERA Data University

Logical Data Modeling Tools – ER/Studio

Organizations looking to improve their data modeling capabilities should evaluate ER/Studio – IDERA's suite of data modeling applications. The tools will help enterprises get the most out of their data resources. IDERA’s ER/Studio provides businesses with the necessary tools to create the conceptual, logical, and physical data models required for efficient database design. ER/Studio is available in three versions that address all stages of data modeling. 

ER/Studio Business Architect

Business Architect is used to create the conceptual data models needed to align business objectives with database design. The tool facilitates mapping the relationships that exist between the people, processes, and data that need to be incorporated into a database. Conceptual models created with Business Architect are easily exported to ER/Studio Data Architect for the creation of logical models. 
Try ER/Studio Business Architect for Free!

ER/Studio Data Architect

ER/Studio Data Architect lets teams take conceptual models and create and manage logical and physical models. The models developed by the tools are suitable for database creation of other data-driven activities like data governance. With Data Architect, teams can create the models required to implement the organization’s business concepts. 
Try ER/Studio Data Architect for Free!

ER/Studio Enterprise Edition

ER/Studio Enterprise Edition helps teams manage and maintain data modeling consistency through the use of a shared model repository. This tool is designed to promote the collaboration that is key to using data models effectively. All stakeholders can be alerted when changes are made to enterprise data models. 
Request a free demo of ER/Studio Enterprise Edition!

Data Modeling Data Governance