Introduction to Data Modeling
Ch.02: DWH | DWH Components | Data Modeling
Video
Data Modeling
Data Modeling Objective
-
Explain what data modeling is and its roles?
-
Be aware of its importance.
-
Explore different types of data modeling.
-
We target to explain the main components and types, and for more details, it could be found in the appendix videos.
What is data model?
-
The data model
-
is An abstract model that organizes elements of data.
-
It describes the objects, entities, and data structure properties, semantic, and constraint.
-
It formalizes the relationship between entities.
-
It describes how the application (report) API data manipulation.
-
It describes the conceptual design of a business or an application with its flow, logic, semantic information (rules), and how things are done.
-
It refers to a set of concepts used in defining such as entities, attributes, relations, or tables.
-
-
Data model is not
-
a science.
-
a static design for each organization.
-
a type of database.
-
a new invention which needs to be done for each project.
-
-
Data model is
-
a general concept that leads to build full architecture.
-
an engineering design practices.
-
different based on the use case and the database type.
-
customizable, and we can utilize some of the ready built architecture.
-
affecting information reporting performance.
-
-
The data model is
-
The first part before starting integration with any new source system.
-
The connection layer between business requirements and technical design.
-
It is also the translation between logical and physical layer.
-
It is unified across all systems and has the same patterns and practices.
-
It engaged with any source systems integration from the early stages.
-
This stage output is a data model design document or mapping sheet
-
Why does the data model are important?
Data models are currently affecting software design.
It decides how engineers think about the problem they are solving.
Data Model Design vs Implementation
-
If you need to build a home, so, how do we design this home?
-
Determine if the home is one level or multi-level and decide main bedrooms and bathrooms for each floor. (User needs)
-
Hire an architect to put the architecture in more detailed way the size for each room, the distribution of the wires, where the plumbing fixtures will be placed, etc. (Architecture phase)
-
Decide the decorations, colors for each room, carpets, etc.
-
-
What do we do for the implementation?
-
Hire a contractor to build (implement the design) the home.
-
This phase implement the design, but it also includes some detail related to the real way to build the tools and the material (Physical Design).
-
Elements of Data Model
Facts: are the measurements/metrics or facts from the business process (Telecom industry, measurement would be the count of daily/hourly usage per customer). We could consider facts as the source of reporting for the business.
Dimensions: provide the context surrounding a business process event. In simple terms, they give who, what, where the fact, (Telecom industry, for the fact daily usage, dimensions would be customer_id, location_id).
Attributes: are the various characteristics of the dimension. In the previous examples, the attributes can be customer details (from customer_id get the gender, age, nationality, etc.).
Elements of Dimensional Data Model
Fact Table: is a primary table in a dimensional model.A Fact Table contains (Measurements/facts and Foreign key to dimension table). It located at the center of a star or snowflake schema and surrounded by dimensions.
Dimension table: contains dimensions of a fact and business reference data. They are joined to fact table via a foreign key. Dimension tables are de-normalized tables. It connected to the fact table and located at the edges of the star or snowflake schema.
Example of Data Model
Elements of Data Model
Dimensional model life cycle:
-
Gathering Requirements (Source Driven, Business/User Driven).
-
Identify granularity of the facts
-
Identify the dimensions
-
Identify the facts
Dimensions Types
-
Conformed Dimension.
-
Degenerate Dimension.
-
Junk Dimension (Garbage Dimension).
-
Role-Playing Dimension.
-
Outrigger Dimension.
-
Snowflake Dimension.
-
Shrunken Rollup Dimension.
-
Swappable Dimension.
-
Slowly changing Dimension.
-
Fast Changing Dimension (Mini Dimension).
-
Heterogenous Dimensions
-
Multi-valued dimensions
Dimensional Modeling Further Reading! 🗒️
Data modeling is an essential part of the data warehouse. So, we gave this part our attention in the course and also seeking to provide some further materials to help the interested practitioners to have more details.
We did a lot of searches, and we found one of the best references in this area is Dimensional Modeling: In a Business Intelligence Environment and you can download the book for free from this link.
Note: We will keep this article updated for all detail for data modeling
Why do we choose this book?
- Simplicity the book is easy for reading for anyone who can dig into this field without complication.
- Practical Examples contains too many examples which help to give more idea about the topics.
- Structure the book is structured in a reasonable manner, which allows you to build the full picture with a quick and logical way with data engineering and data warehouse perspective.
Data and Dimensional Modeling Design Further reading
We recommend reading the following chapters as further reading as it gives you more detail about this topic with more examples.
- Chapter 3: Data modeling: The organizing structure from page 47 to page 73.
- Chapter 5: Dimensional Model Design Life Cycle from page 103 to 120
Assignment: Please write an article explain some parts we missed in the videos or didn't dig in detail, and after reading this part, you released it could be part of our course.
Reference Link
- The data model design vs implementation example taken from: https://opentextbc.ca/dbdesign01/chapter/chapter-5-data-modelling/