Analyses Driven By Normalized Relationships
Many types of data warehouse analysis are driven by the normalized relationships in the database.
There is, for example, a considerable amount of analysis made within the data warehouse that is not quantitative in any way. For example, market basket studies determine clusters of products that are most frequently purchased together. A market basket study does no counting, aggregating, or any other quantitative measure or analysis.
Tracking the cycle times for various entities in a data warehouse requires a sophisticated method of following the relationships between statuses and events. Modeling strategic frameworks such as value chains, supply chain management, competency models, industry structure analysis, and total quality management requires sophisticated relational models for which denormalized approaches are inadequate.