Scribd
Upload
130 views6 pages

Data Warehouse Data Modeling and ETL Designs.

This document provides a quick introduction to data warehouse modeling and ETL design. It discusses the goals of a data warehouse including making analysis queries easy to write. It describes transactional databases and data movements, and how dimensions are used to locate facts. It also discusses star schema modeling and lists some things that should be done for data modeling.

Uploaded by

ThomasMelachridis
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
130 views6 pages

Data Warehouse Data Modeling and ETL Designs.

This document provides a quick introduction to data warehouse modeling and ETL design. It discusses the goals of a data warehouse including making analysis queries easy to write. It describes transactional databases and data movements, and how dimensions are used to locate facts. It also discusses star schema modeling and lists some things that should be done for data modeling.

Uploaded by

ThomasMelachridis
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

A QUICK

INTRODUCTION TO:
DATA WAREHOUSE:
DATA MODELING
AND ETL DESIGNS.
Introduction.

The goals of a data warehouse.

The Real Goal.

Transactional Databases and data


movements.

Locating Facts using Dimensions.

Star Schema.

Another list of things that should be


done for data modeling.

workearly.gr
INTRODUCTION

If you are working data sets with B2B customers, the data sets will be most likely
diverse and make it difficult to perceive all the queries(kinds). To help with this
problem we have data modeling, which can be rather useful while designing ETL
systems.

If you have a business logic goal and design ETL correctly then you can achieve the
major goals of a data warehouse.

A list of things achieved by the ETL processes:

- Showing information consistently.

- Timeliness.

- Adaptability.

workearly.gr
THE GOALS OF A DATA WAREHOUSE.

The DW/BI systems must have the following things:

THE REAL GOAL

The main goal of building a data warehouse is to make it easy for analysts, to write
analysis queries quickly and effectively.

Business and Data Analysts should be able to do the following:

- Descriptive Columns.

- Simple joins.

- Guarantee the uniqueness of data when asked for(SELECT DISTINCT).

- Correct Aggregation.

- Query data for granular and larger levels.

workearly.gr
TRANSACTIONAL DATABASES AND DATA MOVEMENTS.

You must remember that the actual use case of your app or product is transactional
but the analysis that you are going to do is rarely going to be transactional.

You got to aggregate or analyze to discover information about the functioning of the
business processes.

That is why you have to ensure that there is no overlap between two events that you
recorded in your transactional database otherwise the aggregation would be wrong,
irreversibly wrong.

It is truly ideal when the data collected has primary keys. But more often than not this
is not the case. You have to have something to remove redundancy.

This is why we use different dimensions of the data to capture unique rows.

LOCATING FACTS USING DIMENSIONS

< Dimension tables are referred to by some as "descriptive context" the “who, what,
where, when, how, and why” of a facts. >

To query the fact tables and get some business insights, you have to consider the
information provided by the dimension tables.

That's why you must know how to leverage the dimensional modeling and query the
facts.

workearly.gr
STAR SCHEMA

When checking the dimensional modeling, it is more beneficial to de-normalize or


normalize, the database design instead of taking this schema for granted.

In dimensional modeling, in some ways, the fact table is most clearly the
combination of dimensions it carries.

Each fact is identified as the unique intersection of values in each of its dimensions.

While there is often some sort of identifier that can serve as a primary key in a fact
table, standard data warehousing practice creates the primary key as a composite of
all the dimensions it carries.

For any combination of the dimensions, there is exactly one fact record, with measures
that can be aggregated and analyzed.

ANOTHER LIST OF THINGS THAT SHOULD BE DONE FOR DATA


MODELING.

workearly.gr

You might also like