The use of data within an enterprise imparts context and meaning to it, creating information. This data, used in analysis and reports, and as raw material in processes and products, has a cost that must be managed to drive efficiencies and increase net benefits. Describing the data from an information perspective (meaning and usage) provides a common architecture for the enterprise, facilitating improvements in design and reuse across the enterprise and so facilitating efficiency savings, and delivering benefits through insights into new opportunities.
Information architecture is defined and described at a conceptual level (analogous to the functional architecture approach) and this is then aligned with a logical model of the entities and their relationships. Implementation in a physical model incorporating database designs and specifying the physical schema is outside the scope but understanding the information architecture is crucial in decisions concerning its implementation.
In principle:
- Data comprising information that is used in many different domains and processes should be consistently collected and sufficiently up to date for its purpose
- There must be ‘one version of the truth’ - an authoritative source of data is essential
- Information used in analysis and reports should be available in a raw state. Decisions made about what data is made available and in what combinations and forms will severely restrict flexibility - analysis tends to evolve as insights are derived from the data so decisions made during provision may need to be revisited at cost as the analysis teams develop their insights
Above all the information architecture approach is concerned with describing data and designing data structures from the perspective of how they are used in the enterprise. The objective as with other domains of architecture is reusability, composability (use within solutions), and driving efficiencies and benefits.