Data Warehouse| Data Mart & its Types

Data mart(DM)

Data mart usually organized as ID model as star schema made of a fact table & multiple dimensions table.

Definitions :

“smaller local data warehouse called data mart”

“ it is a subset of data warehouse& it support a particular region , business unit, or business function.”

“ it is a collection of subject area organized for decision support based on  the need of a given department”

Data mart is a simple form of data warehouse that :

·         Focus on simple subject

·         Controlled by single department

·         Smaller & less complexity as compare to data warehouse

Types of Data Mart

    1.       Dependent DM

    2.       Independent DM

    3.       Hybrid DM

    4.       Temporal DM

    5.       Non – Temporal  DM

   1)      Dependent data mart : These Data Mart draw data from a central data warehouse that has already been created . In dependent data mart the process of extraction, transformation is simplified because summarized clean data has been loaded into the data warehouse

Dependent data mart are usually build to achieve improved performance and availability of lower communication costs.  

2)      Independent data mart : These data mart draw data directly from operation or external sources of data. In independent data mart the ETL process has to go through number of stages i.e. the process is not simplified. The data is collected from different external sources. Independent data mar t are basically created to have a solution for a particular problem in a short period of time.

3)      Hybrid data mart : such type of data mart contain the features of dependent data mart and independent data mart . in such data mart’s the input is come from both data warehouse as well as ODS(operational data source) . this type of data mart are specially used to handle the integration of data from multiple source. The ETL(external source ) process in hybrid data marat also has to go through a number of steps because datmart is collected contain features of both data warehouse as well as external source

4) Temporal data mart : These are the data mart’s where: 

1. all the history is displayed i.e  each dimension reflects a number of facts and historical hierarchies

2. Dimensions reflects the hierarchies at the time when event took place

3. Are used for  depth trend analysis

4. Too slow for users who want to view the current position

5) Non Temporal data mart: These are the data mart’s where:

1.All the history is not retained

2.Dimensions reflects the current hierarchies

3.Are used for current analysis 

4.Helps in reducing the size of data & simplified the data in dimensions