Principles
Data Availability - Enterprise data will be made readily available, so as not to delay the business processes, and will enable appropriate sharing across the enterprise
Rationale: Data integrity is at its highest level when the central management of changes to data is done by an authoritative source of record.
Data Creation - All enterprise data should be captured once at the point of its creation.
Rationale: This principle will promote the efficiency; accuracy and consistency of data.
Data As a Business Asset - Data is a business asset and will be organized and managed to ensure that its value to the enterprise is maximized.
Rationale: Organizing and managing the key data assets of the company drive the business processes needed to run the enterprise.
Support Federal Mandates - Data supporting any federal government mandate, such as the current eGov initiative, is defined as enterprise data.
Rationale: The data will support the mandatory policies and procedures of any federal government mandates.
Data Ownership - All enterprise data will have an identified business owner and a technical owner. The business owner will be responsible for defining and publishing the logical data, defining the sensitivity and criticality of the data and approving changes from a business perspective. The technical owner will define and publish the physical implementation of the logical data, adhering to the architectural data standards.
Rationale: Lack of a defined business and technical owner will lead to confusion as to who can change the data.
Standardization of Shared Data - Enterprise data standards should be identified when the value of interoperability with other information systems exceeds the value of uniqueness.
Rationale: There is a cost to standardization and a tradeoff analysis should be done.
Standardization of Common Data - Enterprise data standards should be identified when the value of commonality across NIH exceeds the value of uniqueness.
Rationale: Standardization may reduce the duplication of effort and provide improved reporting. It may reduce the number of IC-managed systems.
Data Integrity - Authority to create and maintain the data will reside with those most knowledgeable about the data or those most able to control its accuracy.
Rationale: Those with the most knowledge of the data will have the greatest ability to maintain it accurately.
Primary Data Source - All enterprise data will have an authoritative, official, primary data source that is the location for all Create, Update and Delete actions. All copies of enterprise data will be considered secondary and will not be updated as part of business transactions.
Rationale: In order for enterprise data to be managed effectively, there can be only one primary source for each data element. Otherwise, inconsistent and erroneous data may result.
Data Identifiers - Every object in the enterprise will contain a globally unique identifier. That identifier will be in the form of the Universally Unique Identifier (UUID).1
Rationale: In an integrated environment, it is frequently necessary to retrieve objects (or data) from other more authoritative sources for that information. To do that reliably requires that the requestor be able to make an unambiguous request. Unique identifiers provide an unambiguous name for the desired data.
