On a recent consulting assignment, I compared an enterprise data model to a data warehouse data model to a data governance data model for gaps and overlaps. Although there were many similarities across the three models, one major difference was structure. The enterprise data model was relational and therefore normalized. The data warehouse data model was dimensional containing measures in fact tables surrounded by dimensions for reporting. The data governance model was a glossary of business terms and definitions. The data governance model was “document-based” (as opposed to relational or dimensional). For example, the data governance perspective considered Order the complete order document, containing order lines and products and customer billing and shipping information. The relational data model had Order separated into Order, Order Line, Product, Customer with billing and shipping roles, and so on. The dimensional data model had order measures such as Order Quantity and Order Amount surrounded by dimensions such as Customer and Product.
While going through the data governance terms, many of which were characterized as document-based, an eerie thought kept entering my mind: Are these data governance concepts really NoSQL concepts?Many of the newer NoSQL (meaning non-relational) databases such as MongoDB take a document view of the world. I have been viewing NoSQL databases as purely a different physical structure, but could it be the case that a data governance glossary is the conceptual or logical perspective for a NoSQL physical perspective? Can NoSQL = YesDG (DG short for Data Governance)?
Have you experienced similar connections between NoSQL technology and business concepts?