Candidate Data Marts

Candidate data marts are suggested from the elements contained in the Business Object Models, Business Object Examples and BI Queries. They are discovered in the workshop and documented in the Candidate Data Mart Table. There is a Candidate Data Mart Table produced for each relationship in the Context Model. The table below is the Customer Candidate Data Marts Table.

This table identifies, at a minimum, the candidate fact tables and dimensions contained in the business objects. They are referred to as candidates to reinforce the notion that a design choice has not been made and that some candidates will, in fact, become data marts, while others will go away as more is learned from the queries gathered in the previous exercise.

At the end of examining the relationships between the customer and the enterprise, the workshop group repeats the process by examining the relationships, business objects and BI queries for the Suppliers, Shippers and Employees.

New business objects, data mart fact tables and data mart dimensions are appended to the Candidate Data Marts List as each business relationship is examined.

In the weeks between workshops and after the final workshop, the OLAP Architects organize the business object exhibits, business object data models and other documentation and notes into folders, one for each candidate data mart. There is also one folder containing the context model, relationship diagrams, business object diagrams and candidates list. These folders may or may not be bound and comprise the BI Survey, the formal requirements for the data warehouse.