1 An Information Model for Federal and EPA Enterprise Architecture Brand Niemann Senior Enterprise Architect U.S. EPA August 5, 2008.

1

An Information Model for Federal and EPA Enterprise Architecture

Brand NiemannSenior Enterprise Architect

U.S. EPAAugust 5, 2008

2

Lots of Requests

• EA Assessment Request for Additional Information:– Other than results captured in the recent

E-Gov Results Report, briefly identify any specific results (up to five) you would like the FEA PMO to highlight in your agency EA assessment. These should be directly attributable architectural results externally meaningful. For example, things like establishing reusable information exchanges, or seeing the cost of implementing a widespread information exchange being reduced through tools or other investments, or examples of consolidation of technology, or reuse of services and solutions, or the like.

• Sent April 28, 2008, on behalf of Kshemendra Paul by Thomas J. Smialowicz, OMB.

3

Lots of Terms

• Enterprise, Data, and Information Architectures (see Wikipedia).

• Information Management and Information Model (see Wikipedia).

• IT Infrastructure Line of Business (see OMB/FEA Web Site).

• Information and Technology Management Segment (in process).

• Information Sharing Segment (in process).

4

Lots of Definitions

• Business Reference Model (BRM):– Government Services Delivery:

• Knowledge Creation and Management:– Knowledge Creation and Management involves the

programs and activities in which the Federal Government creates or develops a body or set of knowledge, the manipulation and analysis of which can provide inherent benefits for both the Federal and private sector.

» See OMB/FEA Web Site.

5

Lots of New Work

• Information and Technology Management Segment:– A line of business – that is, a set of activities and

endeavors – whereby EPA manages (plans, acquires, develops, operates, maintains, protects, changes) its information holdings, its information systems, and its information technology infrastructure.

• Systems Development, Lifecycle/Change Management, System Maintenance, IT Infrastructure Management, IT Security, Records Management, and Information Management!

6

Lots of Rethinking

• If it is all about the actual information, and this segment is a Herculean task, we might need to rethink how we go about this so we can actually construct an information model of our information management that we and the public can understand so the information can actually be found, accessed, searched, reused, etc.– When the costs of maintaining the individual information systems

increases to the point where you can no longer afford to manage and share the information across the enterprise then it is probably time to migrate the information to the new Enterprise 2.0/Web 2.0 Infrastructure.

• See my EPA Mindtouch Deki Web 2.0 work as an example of that.

7

Parts of EPA Are Looking

• Vacancy Announcement for GS-15 Information Management Position at EPA:– EPA is looking for a supervisory person to lead a

large and diverse branch that is responsible for development and implementation of information management, access and exploitation tools; geospatial modeling and data analysis; acquisition and management of databases related to ecotoxicology, species attributes and habitats, and pesticide fate and transport; tracking of division risk assessment actions; web development; and management of administrative functions and extramural resources.

• Source: EPA Email, September 27, 2007.

8

Context Setting

• We have to start somewhere to create a context for what we are doing in our own work or in this case, for FEA and EPA information modeling.– Process – from a simple “to do list” to a formal

ontology (e.g. NIST Process Specification Language)

– Knowledgebase – from a simple “table of contents” to a formal ontology (e.g. Vulcan’s Project Halo)

See A Process for Building a Knowledgebase, August 7, 2008, slide 2.

9

Application to FEA and EPA

• Information Modeling for EA, DA, Training, etc.:– I have put all of the EA training materials and my

notes in the Wiki – supports training in 2009.– I have put the Architecture Development Standards

and Guidance document in the Wiki – supports management of deliverables over the next year.

– I have put a lot of Kevin Kirby’s Data Architecture work in several wikis – supports “Metadata in Web 2.0”:

• See Michael Daconta, Government Computer News, March 17, June 16, and July 21, 2008.

– I have put a lot of FEA information in the Wiki – supports transparent connection between EPA and FEA.

See Ken Orr Special Session, August 7, 2008, slide 19.

10

An FEA Information Model• The FEA has an XML Schema for the Exhibit A-300, but that is very

limited as an information model.• Registries and Repositories (e.g. Core.gov) try to do a better job of

managing information, but they are expensive and are not as transparent and collaborative as Web 2.0 Wikis.

• There is also a need to support decision making where Enterprise Architecture leaves off and Solutions Engineering takes over.– Architecture bookkeeping, training, and experience along with

actual solutions development work.• So use the upcoming 2008 Enterprise Architecture Conference:

– It reflects lots of subject matter expertise and practical experience, but with no real structure and relationships (lacks an ontology).

– So start to build a knowledgebase with structure and relationships and continue to drill down with the information to at least one agency like EPA!

11

EA 2008 Conference Process

• Annual Conference Scheduled• Program Advisory Board Appointed (Subject Matter

Experts):– The Program Advisory Board guides and supports the

development of the event content to ensure a relevant, first quality program that meets the educational needs of public sector participants. Board members fulfill this role by participating in periodic meetings, making recommendations about the session topics and speakers, and identifying organizations that would benefit from conference participation.

• Call for Participation Proposals• Select Session Topics and Speakers• Produce a Successful Conference:

– But it should not end here – there should be a knowledgebase built before, during and afterwards.

12

EA 2008 Conference Information Model

• Topic (theme): Using EA to Manage Change and Realize Outcomes

• Subtopics (sub-themes):– General Session– Keynote– Networking Reception– Plenary– Professional Education Program– Town Hall Meeting

13

EA 2008 Conference Information Model

• Subtopics (examples):– General Session:

• An Overview of the Next Generation of EA Practice Guidance

– Keynote:• An Executive Architect’s View of IT Asset Investment and EA

Governance Strategies

– Networking Reception:• Networking Reception with Sponsors

– Plenary:• Putting EA to Work for Your Agency: Getting Started

– Professional Education Program:• Software-as- a-Service and the Future of the Enterprise Architecture

– Town Hall Meeting:• What it Means to be an Architect in the Government Today

14

EA 2008 Conference Information Model

http://govinfo1105.wik.is/Enterprise_Architecture_2008%3a_Using_EA_to_Manage_Change_and_Realize_Outcomes

In Process

15

EA 2008 Conference Information Model

http://govinfo1105.wik.is/Enterprise_Architecture_2008%3a_Using_EA_to_Manage_Change_and_Realize_Outcomes

In Process

16

2008 Government Leadership Summit Information Model

http://govinfo1105.wik.is/2008_Government_Leadership_Summit_June_1-3

An example of a morecomplete information model.

17

2008 Government Leadership Summit Information Model

http://govinfo1105.wik.is/2008_Government_Leadership_Summit_June_1-3

An example of a morecomplete information model.

18

EPA Information Model

• Ontology for Segment Data Architecture:– Ontology - categories of interest in a domain

and the relationships among them (note this is independent of syntax and technology)

• Source: John Sowa, Knowledge Representation, 2000.

– Ontologies – semantic models that are the basis for Semantic Web systems.

• Source: Semantic Web for the Working Ontologist, page 1

19

EPA Information Model

• Semantic Web for the Working Ontologist: Effective Modeling in RDFS and OWL:– Chapter 2: Semantic Modeling, Fundamental

Concepts:• Modeling-Making sense of unorganized information.• Formality/Informality-The degree to which the meaning of a

modeling language is given independent of a particular speaker or audience.

• Commonality and Variability-A fundamental aspect of the Semantic Web that a model can represent.

• Expressivity-The ability of the modeling language to describe certain aspects of the world. More expressive modeling languages can express a wider variety of statements about the model. Modeling languages of the Semantic Web – RDF, RDFS, and OWL – differ in their levels of expressivity.

20

EPA Information Model

• Semantic Web for the Working Ontologist: Effective Modeling in RDFS and OWL:– Chapter 4: Semantic Web Application

Architecture:• Contains and excellent example of the application

of ontology to Enterprise Architecture, specifically, the Federal Enterprise Architecture Reference Model Ontology (FEARMO) (see pages 248-258).

21

EPA Information Model

• Ontology for Segment Data Architecture:– Following the basic process for building an ontology,

we state the Concept, give a Definition, and provide a Specific Example of how the Definition relates to the Concept.

– Then we drill down into the parts of each of the three basic concepts above (e.g. Segments are composed of, etc.) and map them one to another (define and specify the ontological relationships). We will also have to harmonize the data elements and metadata across the Segments, Systems, and Data Models (a big, but necessary job!)

22

EPA Information Model

• EPA Segments for 2008:– Air Quality Management – Water Quality Management – Land Quality Management (see OSWER Segment)

– Emergency Management – Substance Management – Administrative Services – Enforcement and Compliance Assurance – Research and Science – Geospatial Services – IT and Information Management – Records Management

23

EPA Information Model

Source: Lisa Jenkins, Architecting A Segment Across Geographic Locations, June 2008, Slide 14.

24

EPA Information Model

http://gov2.wik.is/EPA_Enterprise_Architecture/Ontology_for_Segment_Data_Architecture

25

EPA Information Model

http://gov2.wik.is/EPA_Enterprise_Architecture/Ontology_for_Segment_Data_Architecture

26

EPA Information Model

http://gov2.wik.is/Ontology_for_Segment_Data_Architecture/Segments/Land_Quality_Management