Possible Futures: Metadata and the HomelandSecurity Mission

1
Possible Futures Metadata and the
Homeland-Security Mission

• Martin Smith, Information Sharing Program
  Manager
  
• Department of Homeland Security
• The Information Management Forum
• Chantilly, VA
  November 20, 2003

2
What Has Changed?

• Technology and other changes open up new ways to
  fulfill business goals
  
• Recent changes (5-10 years)
• The Internet the death of distance (and time
  isnt feeling well)
  
• New business models the agile organization
• Very recent changes (2-4 years)
• Emergence of XML
• Practical component-based architecture
• Terror threat response focused on information
  sharing
  
• About to change
• The real-time organization

3
New Roles for Metadata

• Smart data
• Transactional intelligence
• Documents versus databases
• Beyond search the uses of structure
• Focus on the interface
• Is this my future?

4
Smart Data

• Emergence of XML removes restrictions on
  complexity of data structures
  
• Each fact can carry an unlimited amount of
  context (born-on date, quality indicator,
  special-handling codes, distribution)
  
• Isnt this expensive??
• Users (mainly systems) can interpret context to
  make complex decisions, combining data context,
  local context, business rules, etc.
  
• Practical fool-proofing for self-service
• Reduce drudge work for experts (but top experts
  needed to define best expression of context
  metadata)
  
• Faster and better consistency scalable without
  limit
  

5
Transactional Intelligence

• DHS (mostly) in the business of making lots of
  decisions, quickly millions served
  
• No time to read reports
• Red-yellow-green at the airport
• Not high policy, but just as important
• Decision quality must be higher (more context
  data)
  
• See you in court (maintain decision trail)
• Flexible and fast - - getting the word out and
  changing procedures everywhere, instantly

6
Documents vs. Databases

• Transactional intelligence requires databases
  (collections of facts)
  
• Not, however, your fathers relational tables
• Documents are collections of facts, with context
  and relationships
  
• Consider the form versus the report
• How far can we go without losing value?
• Whats the benefit? . . .(next page)

7
Beyond search value of structure

• Contrarian assertion most search activity is a
  symptom of poor system design or poorly
  understood business process
  
• A workaround thats turned into a goal
• Better search touted as goal of data tagging!
• The data you need should come to you!
• Real value of structure is to model real-world
  meaningful objects, like terrorists
  

8
Focus on the Interface

• The components revolution in
  computer architecture is here
  
• See also service-oriented architecture, OOD
• Benefits are compelling faster, better, cheaper,
  even easier
  
• FEA and DHS are committed to this direction
• The end of applications all you need to know you
  learned in kindergarden, playing with Lego
  blocks
  
• Change your process daily (if the humans can keep
  up!)
  
• So what does this have to do with metadata?
• Components require a standard interface
  contract
  
• The contract consists of agreement on standard
  communications protocols and the exchange of
  standard XML data structures
  
• Managing XML schemas defining exchanges between
  components will be the main business of metadata
  programs

9
Is this my future?

• No!
• Initial reaction of many non-IT people to XML has
  been negative
  
• Fear not analysts will not type , or
  even see them a misunderstanding caused by crude
  first-generation XML tools
  
• Most metadata will be generated automatically
• Things like dates, audit info
• Almost all the rest will come from pick-lists on
  a form instead of free-text typing (hey, fewer
  keystrokes and fewer erros, er, errors!)
  
• But wait, theres more! Metadata-enhanced
  systems will automatically populate your
  research files
  
• Bottom line less drudge work, more quality time

10
The Inevitable Disclaimer
The Metadata Goldmining Magic Quadrant
(with apologies to the Gartner Group)
Ability to
Ability to execute
DHS
Completeness of vision