Business Intelligence

What is Business Intelligence?

The Process of converting raw data into useful information using set of tools which helps an organisation for high level management to take effective decisions to improve the business Processes.

Any BI project includes following two processes:-

1. ETL Process (Extract Transformation Loading)
2. Reporting & Analysis Process

We will see the processes in detail in next section.

Linked Dimension Vs Conformed Dimension

Linked Dimension: This dimension joins different fact tables from two or more cubes

Conformed Dimension: This dimension joins multiple fact tables with in a cube

What is Self-referencing dimensions?

This occurs when a dimension references itself by a primary and a foreign-key relationship in the same dimension.

Example: Dim_A -> Dim_A

What is Fact Dimension?

Sometimes we need to analyze our data based on the some specific value which are available only at fact table. In this case, the particular data will act as a dimension. This dimension is called as Fact Dimension.

Fact Dimension also known as Degenerated Dimension

Fact Dimension will be more helpful, when we need to categorize the data based on similar kinds of data.

Example: Invoice Number, Purchase Order No, Etc.,

Of course, we can maintain separate dimension table instead of using Fact Dimensions, but no benefit, because the dimension table grow in the same ratio of fact table growth. So by creating Fact dimension, we can avoid duplicate data and reducing memory consumption. Also it helps to improve the performance during data retrieval time.

Data modelling Vs Dimensional modelling

Data modelling:
Modelling data storage optimized for Transactional Processing is Data Modelling

Dimensional modelling:
Modelling data storage optimized for Analytical Processing is Dimension Modelling..

SSAS: What is Named Query?

1. Named Query is a SQL Statements and it act as a another table

2. It is formed by selecting rows and columns from single tables or relating multiple tables

3. It helps to make the data source view with new table structure without modifying the underlying data source