Do we need Data Dictionary or not?

Is data dictionary a good choice for developing a database or not. I am having a confusion as I can make database directly without any database schema, but our professor insists on making a one. So I need some reason as why and what is the use of making data dictionary?

When you are considering a database with some 4-5 tables and very less foreign key relations between them then its fine.

But when you come into the IT field and start projects in production and try to design the database with tables ranging from 20-50 or more along with a huge number of cross foreign key relationships then the thing gets really complex.

So its a good practice to use the data dictionary for better working.

I found that data dictionary is used huge in the database application and in the field of machine learning, and i have been using since long.

Lets have some formal definition :

Data dictionary is defined as the set of information which is responsible for describing the contents, format, and structure of a database and the relationship between its elements.

The data dictionary are basically used to control access to and manipulation of the database.

LETS point out some notes:

• They are the the lists of all of the names used in the system models.
• The descriptions of the entities, relationships and attributes are also included here.
• Many CASE workbenches support data dictionaries.

Advantages of Data Dictionary :

• It supports name management and avoid duplication.
• It can be used as the store of organizational knowledge linking analysis, design and implementation.

The data dictionary is very significant since it gives information such as what is in the database, who is allowed to access it, where the physically stored database is, etc. The database users usually don't communicate with the data dictionary, it's operated only by the database administrators.