Why this is in my collection

From the publisher:

A quick and reliable way to build proven databases for core business functions Industry experts raved about The Data Model Resource Book when it was first published in March 1997 because it provided a simple, cost-effective way to design databases for core business functions. Len Silverston has now revised and updated the hugely successful 1st Edition, while adding a companion volume to take care of more specific requirements of different businesses. This updated volume provides a common set of data models for specific core functions shared by most businesses like human resources management, a

Highlights

  • Businesses share structure — Silverston's core claim is that most enterprises need the same underlying data structures for people, products, orders, and agreements; that recurrence is exactly the pattern density MDDE depends on, because what repeats can be generated.
  • The Party pattern as durable abstraction — modeling persons and organizations as one supertype with roles and relationships survives reorganizations that would break a naive customer/supplier/employee split; the abstraction level of the model determines how long it stays true.
  • Patterns as a library, not inspiration — the book is literally a catalog of assemble-ready models, shifting the modeler's job from inventing structures to selecting and adapting proven ones; this is the modeling-side twin of pipeline generation.
  • Generalization is a dial, not a virtue — Silverston presents each model at multiple abstraction levels and makes choosing the level an explicit design decision, the same judgment call ensemble modelers face when deciding how wide a hub's business concept should be.
  • Spend design effort only where the business differs — adopt universal models for the common 80% so scarce modeling capacity goes to the truly differentiating 20%; a resource-allocation rule that generation makes even sharper.
  • Model quality compounds downstream — every structure decision is inherited by every database, interface, and report built on it, which is why investing in the model is the cheapest intervention point in the whole chain.

Highlights on this page are generated with the help of AI.