[ About Info|Ed ] [ TiAC ] [ What's new at Info|Ed and TiAC ] [ TiAC White Paper Repository ] [ Information Architecure Resources ] [ White Paper Request ]

Data Architecture - An Overview

© The Information Architects' Cooperative (TiAC) 1996

Data Architecture

A Data Architecture is nothing more than an organized and structured "map" of the enterprises' data, needed to support the business' objectives in a common and shared fashion. It is the focal point for the development of a complete Information Architecture, consisting of a "data view", "systems view" and a "technology view" managed within an IM Framework. The Data Architecture represents information about the business' data and is of primary interest to IS professionals charged with ensuring that information systems meet the needs of the business. This information about data is often referred to as "metadata"

Data Models

Data models are used to portray graphically the data types and the inter-relationships between them. In addition, the metadata also describes verbally the agreed to definitions of each data type (by all stakeholders), known validation rules and formats, business owners of the data type and any other documentation deemed important by the IM function and line business that warrants being collected and managed in a business-like fashion

The last three lines above are key: …and any other documentation deemed important by the IM function and line business that warrants being collected and managed in a business-like fashion because just developing an architecture is a challenging task, because of the political implications "common and shared definitions" incite. Even more difficult is managing the architecture. Metadata represents information of interest to the IM Community, just like Materials Data is of primary interest to the Material Management Community and Personnel Data is of interest to the HR Community. It needs to be managed across its complete life cycle from creation, usage, modification and possible discarding as business needs for the information change.

Data Dictionary/ Data Reposity

To help manage this "metadata" we need databases just like any other business function that needs to manage and share its information. This database is often referred to as a DATA DICTIONARY if all it manages are data definitions, or a REPOSITORY if it manages additional information of interest to the IM community, such as ownership, systems using it, organizations it serves, etc.

A Data Architecture can exist at different levels of abstraction depending on the views of the information it is communicating or representing, similar to a house drawing. There can be an owner's view (or schematic) identifying major needs but not accurate to allow construction to begin, an architects view ,paying more attention to following agreed to building guidelines and codes as well as aesthetics and finally the contractors view showing exact building specification, material requirements, etc.

A data architecture provides a map of the business' data to:

- enable subsequent and improved data integration and reuse

- provide improved protect scoping

- improved responsiveness to business change

- improved management of systems development sequencing and prioritization

- improved level of business involvement