Federal Enterprise Architecture Reference Model Revision Submission Form (DRM 3.0)

Last modified

Created by Brand Niemann, April 6, 2009, 5:52 p.m.

See http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2009_Communique October 8, 2009 Slides

Architecture & Infrastructure Committee: Governance Subcommittee Meeting (Agenda) NEED Report on Compilation of DRM Comments, Brand Niemann and Rick Murphy, DRM Community of Practice, 
April 10, 2009. SlidesCompilation of Comments.

In terms of my submissions, you may like to review the original language from November which contains sufficient detail to pass this stage of the RMMP screening process. Best wishes, Rick

See David C. Hay Comments, April 15th and two sets of slides at bottom of page.  david.hay@capgemini.com,  dch@essentialstrategies.com

See EPA DRM 3.0 and Data.Gov Pilot, April 21

See A Target Enterprise and Data Architecture for the US EPA: Implementing DRM 3.0 and Data.gov, April 23

See David C. Hay Comments, April 22nd

April 29, 2009, Research on Data Management and Information Sharing Technology in the United States (Japanese Delegation to DoD) led by Professor Hajime Horiuchi, Professor of Information Engineering, Faculty of Commerce,
Tokyo International University and Vice President of UMTP Japan. Invited Briefing. Questionnaire for the DoD MDR and Net Centric Data Strategy.

May 1, 2009, W3C Use Case for Persistent URIs.

1 Introduction

For the Federal Enterprise Architecture (FEA) Reference Models to remain useful and relevant, they must be periodically updated to reflect new concepts, technologies, and priorities. This form is designed for agencies to submit information about proposed revisions to the FEA Reference Models. For more information on the FEA Reference Model maintenance process, please refer to the “Federal Enterprise Architecture Maintenance Process” document on http://www.egov.gov.

2 Instructions

Please complete all sections of the form, including the submitter’s contact information (Submitter Information), summary information about the revision (Revision Summary Information), and any additional information specific to the FEA Reference Model being revised (Reference Model-Specific Information). Please send the completed form to fea@omb.eop.gov for evaluation. The submitter will be contacted to acknowledge receipt of the form and for follow-up on any outstanding questions or issues with the submission.

3 Revision Summary Information


Name Brand Niemann and Rick Murphy
Agency US EPA and GSA
Phone 202-564-9491 and 202-501-9199
E-Mail niemann.brand@epa.gov and richard.murphy@gsa.gov


1. Reference Model (CRM) to be revised:

___ Performance Reference Model (PRM)
___ Business Reference Model (BRM)
___ Service Component Reference Model (SRM)
__X_ Data Reference Model (DRM)
___ Technical Reference Model (TRM)

2. Type of revision:

__X_ Modification
__X_ Addition
___ Deletion

3. Provide a short summary of the proposed revision.

1. By now it should be clear from the President's Open Government Directive and the evolving work on data.gov that Open Government Data is a priority for this administration. Both Vivek Kundra and Beth Noveck are playing leadership roles in this area.
2. The 8 Open Government Data principles present both public policy and technology policy issues that need to be debated thoroughly as part of the Reference Model Maintenance Process (RMMP). The 2005 Joint Proposal which defines the RMMP specifies that the Assessment Team engage subject matter experts when evaluating the submissions. Among others Carl Malamud of http://public.resource.org and Greg Elin of the Sunlight Foundation would serve as excellent subject matter experts.
3. Greg and I already had a telcon with Suzanne Acar and Adrian Gartner. Suzanne and Adrian invited Greg to present at a DAS meeting and Greg is in the process of following up. 
4. We make Open Government Data available on the Internet and adopting W3C's URI scheme as part of the DRM is essential to the most successful implementation in Federal IT systems of the three goals outlined in the President's Open Government Directive: transparency, participation and collaboration. A few weeks ago, Suzanne and I spoke with Jose Alonso at a W3C meeting and Suzanne was also kind enough to invite Jose to present at an upcoming DAS meeting. I also suggest that subject matter experts from W3C are engaged in the RMMP assessment to support my URI submission. I don't think this submission requires as careful debate, but W3C can serve a vital role in providing the right understanding for its adoption.

4. Have the CIO and Chief Architect for your agency reviewed and concurred with the proposed revision? No CIO at EPA and Chief Architect is on OMB Detail, but is aware of these proposed revisions, and Not Sure about GSA CIO and Chief Architect.


1. Explain the relationship between the proposed revision and related existing federal legislation, regulation, or policy.

Open Government Data is a priority for this administration.

2. Identify any existing or planned federal investments, programs, initiatives, or other efforts related to the proposed revision.

Work on http://data.gov

3. Identify existing government or industry governing bodies or communities of practice related to the proposed revision (e.g., Organization for the Advancement of Structured Information Standards (OASIS), National Institute of Standards and Technology (NIST), Federal Identity and Credentialing Committee (FICC), etc.).

Data Architecture Subcommittee, W3C E-Government WG, Sunlight Foundation, IAC/ACT EA and Data Architecture SIGs, etc. See list of contributors to DRM Comments.

4. Identify any government or industry standards supporting the proposed revision.

W3C, etc.

5. Is the proposed revision proprietary to a single vendor?

No - based on open standards.

6. Identify any existing commercially available products or services supporting the proposed revision.

See for example vendors that presented at Semantic_Community-Semantic_Exchange_February_17,_2009


1. What existing problem or opportunity does this proposed revision address?

Open Government Data is a priority for this administration.

2. What new cross-agency insights or collaboration opportunities are possible as a result of adopting the proposed revision?


3. How could the proposed revisions help agencies make better investment decisions?

The public would see the benefits of their investments at http://data.gov.

4. What other benefits could result from adopting the proposed revision?

The public would feel that we really mean to implement Transparency, Openess, and Collaboration.


1. What effects could the proposed revision have on the other FEA Reference Models?


2. Describe the impacts the proposed revision could have on those outside your agency (e.g., other agencies, vendors, state and local governments, businesses, citizens, etc.).

The public would feel that we really mean to implement Transparency, Openess, and Collaboration.

3. Has the proposed revision been discussed with those who could be impacted by the proposed revision? Yes

If so, with whom has the proposed revision been discussed and what were the outcomes of the discussions?

See list in Section 3.3 item 3 above.

4. Describe any outstanding issues needing to be addressed before adopting the proposed revision.

The schedule for formally developing and implementing these suggestions.

4 Reference Model-Specific Information

4.1 PRM

1. How does the proposed revision improve the integration of the FEA with other federal performance management processes (e.g., agency annual performance plans, agency annual program performance reports, Performance Assessment Rating Tool (PART), etc.)?

2. How does the proposed revision enhance performance-based decision-making?

3. How does the proposed revision create a better “line of sight” between inputs, outputs, and outcomes?

4. How does the proposed change create a better “line of sight” between an agency’s mission and its IT investments?

4.2 BRM

1. Is the proposed revision related to an inherently governmental business function?

2. How does the proposed revision affect existing elements of the PRM (e.g., specific Measurement Areas, Categories, Groupings, or Indicators)?

4.3 SRM

1. Explain whether the proposed revision is tightly aligned to a specific BRM line of business or sub-function or is cross-cutting in nature (i.e., could support multiple BRM lines of business or sub-functions).

4.4 DRM

1. How does the proposed change improve the ability of the Federal government to create, collect, categorize, inventory, preserve, disseminate, discover, retrieve, or share data/information?

See Compilation of Federal Data Reference Model 2.0 Comments Received  and David Hay Suggestion for Improved Data Description Catalog .

Note: http://data.gov/ (data available in downloadable formats in late May 2009) and Federal Data (information provided by semantic publishing now)!     

4.5 TRM

1. Explain whether the proposed revision is tightly aligned to a specific SRM Service Domain, Service Type, or Component or is cross-cutting in nature (i.e., could support multiple SRM Service Domains, Service Types, or Components).

2. How does the proposed revision affect existing elements of the SRM (e.g., specific Service Domains, Service Types, or Components)?


David C. Hay 04/22/2009 08:54 AM


Pavithra pavithra_kenjige@yahoo.com

Brand Niemann

Or, should I offer my sympathies.  I am pleased to hear that you are willing to work to promote my conceptual data model patterns.  As I said in the memo to Niemann, I did present them to the Library of Congress last month, and I do believe that they are at a level where they would be generally applicable.  I am on the road at the moment, but later this week I will be happy to send them to you.

Meanwhile, to get an idea of how I approach these things, you may want to get a copy of my book, Data Model Patterns: Conventions of Thought.  I wrote that 15 years ago, and my current patterns are a little more subtle, but the ideas I was conveying there are still true. It also is an example of how I think models should be documented.

That book was an exploration of the concepts, and I pretty much followed my nose in determining how to balance a need for models being general enough to be robust but specific enough for people to understand them.  I got criticisms for being either too general or too concrete, so they must have been just about right.

The book I’m working on—whose models I am submitting to the DAS—addresses the issue of abstraction level head on.  The models that should be of interest to the Federal Government are at what I call “Abstraction Level 1”. These initially cover the general areas of people and organizations, geography, physical assets, and activities.  There are two compound areas:  “Facility” which is a place with a purpose, where people and organizations make use of physical assets to perform activities; and “Contract”, where people and organizations obligate themselves to provide physical assets and activities to other people and organizations.

There is also an “abstraction level 0”, which is a pattern behind the first four areas.  Abstraction Level 2 is standard models for standard functions, like human resources, laboratory, etc.

“Abstraction Level 3” is about specific industries.  These are not complete models for such a company (or agency), but only that which makes it unique.  Pretty much any company or agency can use about 80% of the level 1 model.  It is the other 20% that make the company or agency interesting.  For Alberta Transportation, for example, it was the data model describing a highway interchange.

In addition to the models, I have written a series of articles about the relationship between data modeling and OWL, that you can find on TDAN.com.  Specifically, “Data Modeling, RDF, and OWL” can be found at:

One: Introduction http://www.tdan.com/view-articles/5025

Two: Entity Classes to OWL Classes http://www.tdan.com/view-articles/5001

Three: Relationships to OWL http://www.tdan.com/view-articles/4594

I will be sending you the models later in the week.  I look forward to working with you. Dave Hay

David C. Hay 04/15/2009 01:42 AM


RE: Presentation of DRM Suggestions on April 10th and Invitation to April 28th Activity

Brand, thank you for your detailed response. I appreciate it and I welcome the opportunity to discuss these things. I do hope that I can make some small contribution.

Point: Wandering off into a direction that will not be successful

Counter-point: Augmenting each of DRM 2.0's three parts (description, context, and sharing) as requested

I understand..

Point: A very good example of an ontology is a conceptual data model, that indeed describes the semantics of an organization

Counter-point: Agree


Point: Coming up with a common ontology to describe the concerns of the federal government, we are talking about data description, not data context

Counter-point: Both

The ontology itself is descriptive.  Connecting the data elements with agency functions puts them in context.

Point: Web site describing the statistical abstract of the United States, but I couldn't tell the relevance

Counter-point: The example at the Web site is similar to your Three Abstraction Levels, namely 1: The Enterprise Model; 2: More Concrete; and 3: Industry Specific. The Statistical Abstract's Four Levels, namely 1: Topic (e.g. Population); 2 Subtopic (e.g. Decennial Censuses); (3) Data Tables (e.g. Table 1. Population and Area) and (4) Data Elements (e.g. Resident Population)

Not quite. What you describe is a classification scheme (dare I say “taxonomy”?) for a body of knowledge. In my case, I am describing essentially the same world in progressively more abstract language. Each presentation represents a particular view of the world.  Some elements are shared across views, but the observer is looking only at one view at a time.

Point: Start with a taxonomy to be approved by the Data Architecture Subcommittee, but that wasn't presented as though it would really happen

Counter-point: Ask Suzanne Acar and Adrian Walker

I would be curious to hear their take.  My point is that the world is much more complex than the hierarchical view of a taxonomy.  Libraries got away with it because a physical book could only be in one place in the library. The cataloguer of a book on music derived from Arabian mathematics had to make a decision as to where in the Dewey Decimal System it fit. 

Point: “Data description” has been relegated to being concerned solely with URIs

Counter-point: Not our intent. The example provides self-describing metadata (e.g.


Point: I believe the data description part of the DRM is the most important

Counter-point: DRM 2.0 thought that all three parts were important

Fair enough.  It’s just my prejudice.

Point: We identify the most important things of significance to the federal government and create a conceptual (that is to say, semantic ) data model to describe them

Counter-point: DRM 2.0 thought that we should not do that. They did a conceptual model of the relationships between the three parts.

Don’t I know that…

Point: "Closed world assumption" versus "open world assumption"

Counter-point: RDBS uses former and Semantic Web uses latter

Point: In the context of DRM, agencies are looking for a data model that can be used as the basis for the architecture of databases, not as the basis for exploring large bodies of unstructured data.  The ontology we want is designed to describe the structures that we want to control data, not discover them

Counter-point: DRM 2.0 left to the agencies to decide what they needed (data models, discovery, etc.)

True.  It is for the agencies to adopt whatever technology and approach that they want.  And it would be great if the DAS could support all the interesting things going on in the semantic world.  It seems to me, though that most of us are data modeling/database custodians of the corporate resource that is data.  That argues for the closed world assumption being helped first.  But that decision is way beyond the scope of this discussion.

Point: A fundamental data model is currently missing from the DRM, and to add it would be a profound enhancement - OK, I am prejudiced.  I have in fact created just such a model

Counter-point: DRM 2.0 decided not to have one and let individuals, agencies, etc. develop them

Again, I know.  I just think that that is unfortunate. An opportunity to provide a real service was missed. (Again, my prejudice.)

Point: The model consists of (1) “atomic” subject areas: e.g. People and organizations; (2) two “composite” subject areas: Facilities and Contracts; and (3) two “meta” models, that connect with all of the rest of the model: Information resources and Accounting. Take over my model and shepherd it through the approval process

Counter-point: Okay, lets find someone/some group to do that

Thanks for listening. I really appreciate it. Dave

Page statistics
2334 view(s) and 9 edit(s)
Social share
Share this page?


This page has no custom tags.
This page has no classifications.