A "Community of Interest" Approach to Data Interoperability

Existing data administration policy in the DoD has produced good results in some pockets, but has not delivered data interoperability for the enterprise as a whole. In this paper we present a concept of operations for a different approach to data interoperability, one based around the notion of "communities of interest". This approach corrects the shortcomings of the current regime in three ways: First, it does not assume a single monolithic data standard. Instead, we work toward data interoperability within separate communities of interest. These groups serve a knowledge management purpose: they extract the shared knowledge necessary for data interoperability from the larger community, then make this knowledge explicit, and finally help to transfer that knowledge back into the whole community. Second, our approach captures knowledge at more than one level of abstraction. Our approach will produce higher-level definitions that can support interoperability in architecture descriptions and information retrieval "semantic tags" in addition to the current implementation-level data standards. Finally, our approach can become part of the system acquisition process, giving both better incentives for program offices to participate, and better metrics for measuring their results.