MULTI DIMENSION MODEL

MULTIDIMENSIONAL MODEL:

We have 4 methods to organize data using multidimensional model,

1. STAR SCHEMA

2. EXTENDED STAR SCHEMA

3. SNOW FLEX SCHEMA

4. HYBRID

To construct data in the data warehouse using multidimensional model, we have 2 approaches

1.      TOP DOWN

2.      BOTTOM UP

Always in the SAP BW we used top down only.

1. STAR SCHEMA:

                        Multi dimension model consist of fact table surrounded by related dimensions or dimension tables.

Fact table always consist of data where as surrounded dimensions consist of master data.

In general this model looks like a star this is the reason this multidimensional model we are calling as star schema, and the physical formula of schema in the form of cube. 

Limitations of star schema:

1.   The master data in the dimension is specific to cube when we required to same master data across different cubes we required to replicate which leads to data management problem 

 

2.   It does not support Multi languages

  

3.   Indexes are saved in the alpha numeric.

2. EXTENDED STAR SCHEMA:

                                                                                                                                                         

                  Extended star scheme is an extension to existing star schema, in which instead of maintaining master data in dimension table we maintain outside, in the form of master data in attribute table, language in inf ‘ in text table,  hierarchy is of dimensions  in hierarchy table and we define the relationship between these tables with structure of dimension table by using a separate table called Surrogate ID (SID) table.

The physical form of this Extended Star Schema will be maintaining using Info cube.

What is Info cube, facts, fact table, dimension, dimension table, SID table and Info objects? We will see in the next session.