Data warehouse model development methodology

Data warehouse model development methodology (for future reference)

 

Based on the business data model, the data model of the data warehouse system is formed through the eight-step conversion process:

Procedure	Action	Target	Description
1	Select data of interest	Determine the scope of inclusion, reduce loading time, and reduce storage requirements	Determine the data elements to be included in the model and consider archiving other data that may be used in the future.
2	Add time in the key	Provide historical data	Add a time component to the key and resolve the result changes in the relationship between the model from "Time Point" to "time period ".
3	Add derived data	Ensures business consistency and improves data delivery performance	Compute and store frequently used or data requiring consistency Algorithms
4	Determine the granularity level	Ensure that the data warehouse is at the correct level of detail	Determine the desired level of detail to balance business needs, performance, and hidden costs
5	Summary data	Simplified data delivery	Collect data based on the use in the data mart
6	Merge objects	Improve data delivery performance	If frequently used data has the same key and a common insertion mode, merge them into an object.
7	Create an array	Improve data delivery performance	Create an array in the attribute entity field when appropriate conditions are met
8	Separate data	Balance Data Acquisition performance and data delivery performance by separating entities	Determines the insertion mode. If the query performance is not significantly reduced, data is detached.