Derived XML Registry

1
Derived XML Registry Prototype

• Started as Exercise in Developing a Methodology
  for Automatically Populating an XML Registry
• Transformed to Journey of Discovery
• Differing Practices, Policies, Portals Make
  Automatic Registry Construction Infeasible Now.
• Practical Requirements for XML Registry Being
  determined by Forces Driving Technological
  Progress
• Rate of Standards Adoption (CAP, Web Services
  Security, WSRP, etc.)
• 9/11, Social/Political/Economic Factors

Original Idea
Conclusion
2
Derived XML Registry Prototype Options - the What

• Starting with Conclusions
• Automated Harvesting of Existing Government
  Documents/Standards Preferred, but Advance
  Preparation Required.
• Search, Crawl Standards Needed.
• WSRP Adoption Encourages Reuse, Repurpose of
  Content Standards Government-wide.
• Establish Service-Oriented Architecture
  Components as Policy Priority.
• Adapting to Portal Structures Required.
• Use both UDDI and ebXML Registries to Accommodate
  Differing Business Cases.

3
Derived XML Registry Prototype Options - the Why

• Problem Review
• Proliferation of Narrowly-Focused Standards,
  Policies, Legislation, Registries, Portals, etc.
• Adoption Uneven Across Federal Government
  Departments Agencies.
• Reaction TO vs. Setting Requirements FOR IT
  Standards!!

4
Derived XML Registry Prototype Options - the Show

• Brief Notes on this Presentation
• Basics Bear Repeating for Maintaining Continuity
  Keeping Related Issues of Legislation Market
  Forces Grouped.
• Slides of Screen Shots added as Examples of
  Specific Points.
• Emphasis and New Additions Highlighted.

5
XML XML-based StandardsOffer Solution, but No
Panacea

• XML Salvages Legacy Databases, Applications,
  Crosses Platforms, but
• XML Schema Standard is Complex.
• Inconsistent Validation Remains.
• Work-arounds for Toolsets cause Backward
  Compatibility Issues when Standards are Refined
  and Toolsets Improved.
• Lack of Best Practices Risks of Duplicate Terms
  in Vocabularies Possible.
• Adoption Drives Better, Faster Improvement
  because...
• Vendors, Standards Builders Respond to Market.

6
Needs Drive Rush for Domain-Specific Standards

• Similar Concerns Breed Similar Standards.
• Overlap Scope Creep Inevitable.
• Specialization Standards Can be Separate from the
  Generalized Standards in the Same Domain, such as
  Optometry in Healthcare.
• Key Terms can have many Standards, such as
  Person.

7
Government Cross Department, Cross Agency Effort
Growing Faster Now

• Presidential Management Agenda for
  Citizen-Centric Government Emphasizes
  Importance of E-Government, Continues Second
  Decade of IT Efficiency Effort.
• Transition Underway from Early Adoption to
  Leadership in Open Standards Adoption Post Y2K.

8
Reminder Legislation, Policy Mandates Continue...

• Government Performance Results Act-1993.
• Requires Strategic Plan with Periodic Reviews.
• Requires Outcome-Related Goals Objectives.
• Government Paperwork Reduction Act-1995.
• Office of Information Regulatory Affairs (OIRA)
  in OMB Reviews ALL Governmental Agency
  Information Collection (Note These Reports are
  the largest number of 29 Total Items Returned
  from Advanced Search of FirstGov.gov).
• Requires ONLY Information Collection Necessary
  for Proper Practical Utility for Agency.

9
Interrelated Legislation, Policy Mandates
Continue...

• Clinger-Cohen Act-1995.
• Information Technology Management Reform Act
  Repeals 1949 Brooks Act Section 111-- GSA No
  Longer Responsible for Government-Wide Automated
  Data Processing Purchasing.
• Requires Agencies to use Capital Planning
  Investment Control Policies (GPIC).
• Government Paperwork Elimination Act-1998.
• Allows Electronic Information Submission
  Transmission in Government Agency Interactions.
• Agencies Allowed Electronic Records Maintenance.

10
Interrelated Legislation, Policy Mandates
Continue...

• Federal Enterprise Architecture (FEA)
• Clinger-Cohen-Based FEA Framework Formed-1999-CIO
  Council (CIOC), Program Management Office
  (FEAPMO)-2002 under OMB, Returning to CIOC.
• Zachman-Based, Promotes Interoperabilty, Sharing
  Common Federal Processes, Information.
• Aids Capital Planning Investment Control
  Policies (CPIC).
• Common Federal Reference Models Provided-2002
• Performance Reference Model (PRM)
• Business Reference Model (BRM)
• Service Component Reference Model (SRM)
• Data Reference Model (DRM) (Note Changes)
• Technical Reference Model (TRM)

11
FEAF Relies on Evolving Data Reference Model

• Recent Developments are Significant
• Core Data Types Focus Group Evolving into Two
  more Narrow Working Groups.
• Emergency Data Exchange Language Distribution
  Element Paves Way for Use-Specific Messaging
• Emergency Content Types Lead .
• Implications for Intergovernmental Communications
  and IT Operations Require Consideration.
• Content-Based Messaging Provides Sensible
  Mechanism for Service-Oriented Architecture
  Components.

12
Interrelated Legislation, Policy Mandates
Continue...

• OMB Circular A-11 Part 7 - 2002
• Planning, Budgeting, Acquisition, and Management
  of Capital Assets for Reaching Strategic
  Results-Based Goals at Lowest Life-Cycle Cost and
  Least Risk.
• Section 300 Requires
• CPIC FEA-Based Documentation of Business Case
  Justification for IT Investments.
• Performance Measurements.
• Using FEA Reference Models in Relation to
  Specific Lines of Business (Implied. Not
  Explicit).
• Identifying Cross-Agency Resources to Share.

13
Interrelated Legislation, Policy Mandates
Continue...

• OMB Identified Lines of Business-2004
• Lines of Business for CPIC FEA-Based Reports
• 2004-2005
• Financial Management.
• Data Statistics.
• Human Resources.
• Monetary Benefits.
• Criminal Investigations.
• Public Health Monitoring.
• 2006
• Financial Management.
• Human Resources Management.
• Grants Management.
• Federal Health Architecture.
• Case Management.

14
Interrelated Legislation, Policy Mandates
Continue...

• OMB Circular A-119
• Federal Participation in the Development Use of
  Voluntary Consensus Standards in Conformity
  Assessment Activities.
• National Technology Transfer Advancement
  Act-1995
• NIST Coordinates Conformity Assessment.
• OMB Revised Circular Directing Agencies to use
  Voluntary Consensus Standards in lieu of
  Government-Unique Standards Except Where
  Inconsistent with Law or Otherwise Impractical.
• Purpose Reduce to Minimum Government Reliance on
  Government-Unique Standards.

15
Legislation Policy Results

• Well-Intended, very Smart in some Specifics,
  Overly Specific in others.
• Creates Confusion-What to do First?
• Duplicated Provisions, Restated Objectives,
  Unstated Implications.
• Specifies Process and Outcomes such as Sharing
  Resources among Agencies without Clear Priorities
  Criteria.
• Overall Normal, Improving.

16
The Plan XML Registry of Existing Standards
Specs

• Using Firstgov.gov
• Evaluate Methods to Automate or Assist in
  Populating an XML Registry of Governmentally-Used
  or -Related Documents, Standards.
• Early Results Indicate that Crawling the
  FirstGov.gov Portal Raises Problem of Crawling a
  Search/Crawl-based Portal.
• Secondary Crawls of Sites Searched by
  FirstGov.govs Search Engine Requires Human Time
  Skill.
• Basic Inefficiency Revealed.

17
The Plan XML Registry

• Using Firstgov.gov
• Portals Fact of Life Use it for its Advantages.
• Architecture is Registry-Ready!

18
The Plan XML Registry Portal

• Using Portal Models
• Portals Fact of Life Make the Best of Good
  Deal.
• For Scalability-Use Existing Models.

19
The Plan XML Registry Portal

• Structure Portal to Serve Communities
• Communities Allow Structuring Resources by
  Agency, Department, ANY Community of Practice.

20
The Plan XML Registry Portal
21
The Plan XML Registry Portal

• Using GovBenefits.gov
• Crawling a Site May Require Iterations over
  Different Criteria to Compare Results, Reach
  Conclusions.

22
The Plan XML Registry One SIZE Fits NONE

• Using Firstgov.gov
• 1- Evaluate Making the Registry Selectively
  Available using ebXML Registry and 2- eXtensible
  Access Control Markup Language (XACML) to
  Restrict Access by Identity Authentication and
  Role-Based Permissions.
• 1-Developing Results Preempted by Apparent
  Requirement for Using Both UDDI and ebXML
  Registry/Repository Standards Based on Respective
  Shares of the Combined Overlapping Public
  Private Markets Served by Government.

23
The Plan XML Registry One SIZE Fits NONE

• Unbiased Means for Differentiating Best Practices
  Between UDDI and ebXML Registry/Repository
  Needed.
• 1-Suggestion to both OASIS TCs Meet to
  Differentiate Respective Scopes Avoid Further
  Scope Creep Beyond Registry-Specific
  Functionalities.

24
The Plan XML Registry Match the Registry to the
Need

• Using Firstgov.gov
• 1- Evaluate Making the Registry Selectively
  Available using ebXML Registry and 2- eXtensible
  Access Control Markup Language (XACML) to
  Restrict Access by Identity Authentication and
  Role-Based Permissions.
• 2-For those Agencies Programs Where ebXML
  Registry/Repository Favored, XACML Role-Based
  Permissions Usable.
• Similar Available Security Measures Dont Require
  ebXML Registry/Repository, Dont Conflict with
  it, (Yet?).

25
The Plan XML Registry Plan Now to Avoid
Lock-In

• Using Firstgov.gov
• Evaluate the Extent of the Task of Using a
  Prototype Process to Develop a Centralized and
  Federated XML Standards Registry to Support the
  IT Investment Tasks of Governmental Agencies.
• Early Results Indicate Task More Complex than
  Estimated.
• Potential Confusion Greater, Therefore the Need
  is More Urgent to Stem Tide of Locked-In
  Solutions, Duplicated Efforts, Standard
  Incompatibility.

26
The Plan XML Registry WSRP Vendor/Platform
Agnostic

• Using Firstgov.gov
• Method 1 WSRP Standard- Plumtree Web Crawler
  Portlet Makes Results Available by Web Browser in
  Humanmarkup.org Public Service Portal.
• Method 1 Result Not Feasible at this Stage Due
  to Unanticipated Human Time Skill Required for
  Multiple Individual Secondary Crawls to Produce
  Concrete Results However, WSRP Still
  Well-Grounded, Sound Choice for Distribution
  Availability.

27
The Plan XML Registry Need Search Standards Now

• Using Firstgov.gov
• Method 2 Use OWL to Search site by Ontological
  Rules to identify and sort .xml and .xsd files by
  Relationships to Criteria Keywords or Phrases,
  such as Healthcare Services or Emergency Event
  Types.
• Further Research Reveals Lack of Search Standards
  May be Problem.
• Recommend Effort to Develop Search Standard(s),
  Crawl Standards Allowing for Standard Metadata
  Capable of Sorting, Indexing According to OWL
  Description Logics.

28
The Plan XML Registry Hypothesis Check

• Using Firstgov.gov
• Method 3 Use Strawman Search Standard(s) to
  Discover Requirements for an XML-Specific Search
  Standard from this Registry Use-Case to be used
  in a Prototype Government XML Standards Registry.
• Early Results Indicate OWL-Based Metadata Offer
  Best Path Toward Developing Prototype Government
  XML Standards Registry Providing Means to Create
  Use-Specific Inventory of Applicable or
  Appropriate Standards, Resources, etc.

29
The Plan XML Registry Hypothesis Check

• Early Results Indicate OWL-Based Metadata
  Possible with Planning and Sufficient Human
  Decision-Making.

30
The Plan XML Registry Hypothesis Check
31
The Plan XML Registry Hypothesis Revision

• Using Firstgov.gov How many which documents?
• Are or Refer to XML?
• Are or Refer to XSD?
• Wrong Questions
• Firstgov.gov apparently uses a Crawler in its
  Search Engine to Find Resources Based on Specific
  Criteria so XML and/or XSD Results Spotty,
  Require Further Crawls.
• Developing Government-wide Agreements on Best
  Practices and Standards is a Better Approach, but
  Continue Testing this Methodology to Document and
  Distribute Recommendations.

32
The Plan XML Registry Hypothesis Revision

• Using Standards.gov How many which
  Sites/Resources Revealed by Search Contain the
  Documents Required?
• Then Crawl Those Sites for Documents that
• Are or Refer to XML?
• Are or Refer to XSD?
• Identify, Sort, Index.
• XSDs that are Standards.
• XSDs that are Agency- or Company-Specific.
• Extract XSD Elements, Attributes, complexTypes,
  simpleTypes, etc.
• Check for Duplicate Terms Definitions-Compare
  and Contrast.

33
The Plan XML Registry Hypothesis Revision

• Using GJXDM Using Models that Make Sense

34
A Start Remains to be Made Derived XML Registry

• What does Our Search Tell Us?
• Apply Ontological Categorization
• Make Ontology of Lines of Business.
• Make Ontology of Cabinet Departments and their
  constituent Agencies.
• Search by Ontological Criteria for Terms.
• Identify Similar or Identical Terms.
• Identify Number of Occurrences of Terms.
• Identify Departments Agencies using Specific
  Terms.

35
Derived XML Registry Conclusions Revisited

• Starting with Conclusions
• Automated Harvesting of Existing Government
  Documents/Standards Preferred, but Advance
  Preparation Required.
• Search, Crawl Standards Needed.
• WSRP Adoption Encourages Reuse, Repurpose of
  Content Standards Government-wide.
• Establish Service-Oriented Architecture
  Components as Policy Priority.
• Adapting to Portal Structures Required.
• Use both UDDI and ebXML Registries to Accommodate
  Differing Business Cases.