Data Warehouse Methodology
The bulleted items below summarize the five major steps in creating a data warehouse. Although every project possesses some distinctive characteristics, and we reflect those considerations in the project plan, our basic methodology remains consistent.
- Meet with client personnel to identify immediate and long–term data requirements.
- Produce a Requirements Specification for the client’s approval.
- Assemble a Design Specification, beginning with recommendations for the necessary hardware and Business Intelligence software. The Design Specification also includes a data model and mock–ups of the key user interface screens. Work with the client’s project team to refine the design for client approval.
- Develop, test, and document the software described in the Design Specification. Throughout this process, transfer knowledge about system support and long–term maintenance to the client’s IT staff.
- Prior to the production roll–out of the data warehouse, provide “train the trainer” instruction for end users of the system.
There is a range of other data warehouse concepts and methodologies including:<?p>
- Logical data model
- Conceptual data model
- Physical data model
- Star Schema
- Data mining
- Staging area
- Multi-dimensional analysis
- Snowflake schema
- OLAP tools