Feature

Digg This!

Page 1 of 2   next page »

The Department of Defense (DoD) Discovery Metadata Specification (DDMS) describes the DOD's preferred approach for decorating data assets with metadata. By providing a common convention for metadata, the DoD is building a common system for asset discovery, search, description, consumption, and security. This article provides a summary of the DDMS's purpose, structure, and capability. Upon completion the reader should have a basic understanding of the DDMS and should know where to go to get more detail and related materials. All questions regarding this article should be directed to Michael Sick at mike@serenesoftware.com.

Metadata

Metadata is typically defined as "data about data." A metadata specification is an attempt to describe the format and content of a metadata convention. Establishing unambiguous conventions for the expression of metadata allows Communities of Interest (COIs) to better leverage their data assets, by making them more visible. Metadata conventions are beneficial because they can aid in the following areas shown in Table 1.

The DDMS establishes a broad set of categories for its metadata as well as a common set of data elements within the larger categories. While the DDMS provides several options for the metadata format (text, HTML, and XML), the formats are clear enough to support consistent metadata production, consumption, and validation. For an alternate look at metadata specifications, see BEA & IBM's ongoing efforts to establish a metadata specification for business computing at: http://dev2dev.bea.com/pub/a/2004/12/emd.html.

DDMS Overview

The United States Department of Defense governs the Army, Air Force, Navy, and a host of supporting intelligence and logistics agencies. Collectively the supporting IT organizations comprise one of the largest, if not the largest, IT enterprises in the world. The DoD's systems are highly variable in their implementation details and the DoD requires a comprehensive approach to making its data assets visible.

In May 2003, the DoD published the DoD Net-Centric Data Strategy that broadly defines the goals and approaches for making data assets available. The DDMS is the official response to the DoD's network-centric view of data discovery and descriptions and is designed to work across varying data formats, types, locations, and classifications. The DDMS specification is governed by the Global Information Grid (GIG) Enterprise Services Metadata Working Group (GES-MWG). The GES-MWG is charged with evolving the specification to continually match the concepts of operations adopted by DoD.

Granularity

The DDMS specification is currently aimed at the higher-level data assets. The DDMS specification does not require that assets be described at the lower "record set" levels. However, the specification is flexible enough to allow tags to be placed at lower levels in the data hierarchy, if the implementers choose to do so.

Structure

DDMS content is separated into Core and Extensible layers. The Core Layer has four predefined element categories: Security, Resource, Summary Content and Format. Each element is assigned an obligation level (Mandatory, Mandatory Unless Not Applicable, Conditional, and Optional). The Extensible Layer is provided to contain content for domain-specific content areas. Additional obligation sets can be expressed through data requirement languages such as XML Schema and additional schemas can be registered in the DoD Metadata Registry.

The core layer is separated into four distinct category sets, Security, Resource, Summary, and Format, each containing elements supporting its designated role. The Security Set provides security-related information intended to classify the document. An external access control system can consume this information and authorize a client to view some or all of the DDMS description or underlying data. The Resource Set contains elements that provide ways to describe administration, maintenance, and pedigree of the data asset. The Summary Content Set describes elements most often associated with data searches and contains elements such as subject, title, and description. The Format Set describes physical aspects of the underlying data such as mime-type. Table 2 shows the Primary Category Elements for the Core Layer.

While the Security elements do not actively protect the data in the DDMS document, they do provide enough information for an external security system to make a decision on what information to provide based on the client's role and credentials. The samples provided in Listings 1-4 were derived from the specification and are intend to give the reader a feel for what a DDMS document might look like.

Page 1 of 2   next page »