Metadata Management in Practice

1
Metadata Management in Practice

• New Directions in Description Series
• March 2006
• Lynn Herbert
• Innovation, Information and Technology Branch
• Service Canada

2
Overview

• Metadata as Key IM Enabler
• Metadata Management Business Case
• Aspects of Metadata Management
• Metadata Framework
• Tools
• Quality of Metatagging
• Change Management Processes
• Vertical and Horizontal Metadata (GC enterprise
  governance and initiatives)
• Current and Future Challenges

3
Metadata as Key IM Enabler

• Metadata more than descriptive enriched
  metadata sets used mostly for delivery and
  life-cycle administration
• Metadata for Delivery
• Dynamic web sites and linking pages, without dead
  links
• Portals and personalization
• Facet combinations to target specific client
  segments and jurisdictions
• English / French pages
• Page sensitive contact and help information
• Advanced and structured search
• Weighting control and sorting
• Addresses sustainability issues for GOL clusters

4
Metadata as Key IM Enabler

• Metadata for Delivery, continued
• Store content once, publish to any number of
  views
• Separate policy and program information from
  service information
• Multi channel use
• Adapt to changing client needs and departmental
  priorities
• Metadata for Administration
• Structured authoring templates
• Re-usable content objects and documents
• Documents management
• Records management and archiving
• Publishing process management
• Business processes and workflows
• Quality control and accountability
• End-to-end lifecycle management of the information

5
Business Case for Metadata Management

• Complex frameworks and schemes are required to
  support all aspects of delivery and
  administration
• Requires governance, tools and processes for
  creating metadata and managing change
• Sustainability
• Demonstrate the value of metadata, and recognize
  that the effort required needs to continue
  providing value (ROI)
• Adaptability
• Continuous improvement, changing business
  priorities and services
• Evolving technologies, vocabularies, concepts,
  information architectures, business processes and
  requirements
• Quality of metadata values
• Harmonization with evolving GC standards

6
Aspects of metadata management

• Metadata Frameworks
• Tools
• Quality of Metatagging
• Change Management Processes
• Vertical and Horizontal Metadata (GC enterprise
  governance and initiatives)

7
Metadata Frameworks

• Framework is first step in managing metadata
• Metadata Element Sets
• Metatag name (e.g. dctitle, dcsubject,
  gcterms.topicGC, etc)
• Label (Title, Subject, Topic, etc)
• Definition (intended scope or purpose of the
  metadata element)
• Datatype
• Metadata Application Profiles (MAP) is a
  standardized set of metadata elements, plus rules
  for applying them
• Comprised of elements from one or more element
  sets
• Inherits rules of the element set plus can
  further define, narrow or enforce (rules define
  how the metatag value is populated and used
  within an application)

8
Metadata Frameworks

• MAPs, continued
• Single or multiple values
• Optional, Mandatory, Conditional
• Schemes and vocabularies to use
• Relationship to other metadata elements and to
  business process
• System auto-population
• Schemes controlled vocabularies or date
  standards, e.g.
• Namespaces for element sets and schemes
• Metadata Registries

9
Tools

• Technical tools are required to manage metadata
  frameworks, assist metatagging, index the
  metadata, and search on metadata values
• Metatagging interfaces for authors
• Elements encoded in user-friendly interfaces,
  using labels
• Populate values for elements and perform
  automated metatagging
• Mandatory metatagging enforced through workflow,
  with business processes for review and approval
• Free-text entry (summary, title)
• Pick list selections (small flat lists)
• Search and browse complex hierarchical and
  relational vocabularies (CST, Location)
• Authors dont have to deal with complexity of
  underlying metadata

10
Tools

• Vocabulary (scheme) management tool
• Administrator interface to create and edit
  controlled vocabularies in both official
  languages
• Governance on vocabulary control centralized and
  enforced
• Ensure vocabulary updating is part of overall
  change management process
• Metadata indexes and searching
• Structured indexes and query language
• Filtered results based upon facet combinations
• Queries can be pre-run for faster results (Verity
  Topics)
• Combination arrays of boolean queries can be
  pre-processed or live
• Embedded within URLs and applications
• Menu choices within pages
• Enhance free-text search results and assist
  weighting and sorting
• Index brokering (federation)

11
Quality of Metatagging

• Issues with metadata value selections
• too many, not enough or too much the same
• confusion between elements or about purpose of
  element
• failure to review system-generated metadata (e.g.
  summary)
• non-expressive titles and descriptions
• managing expectations both negative and
  positive

12
Quality of Metatagging

• Some solutions
• Workflow procedures for metatagging review and
  revision by experts, without creating delays
• Ongoing metatagging documentation and training
  support
• Continuous demonstration of the purpose and value
  of metadata - showing authors the why of
  metadata is most effective, especially in dynamic
  site areas or when metadata is used in enhanced
  search
• Use technology to limit number of choices
• Construct metadata reports
• English/French consistency
• Tracking dynamic page performance
• Information architecture and taxonomy usage and
  review
• Usage reports by element
• Monitor quality of descriptions and titles

13
Change Management Processes

• In a fully functional metadata system changes may
  have little impact on the business authors or
  metatag specialists, but need to be well planned
  to ensure the changes are incorporated at all
  points in the system
• Framework registry and documentation
• Interfaces
• Vocabularies
• Workflows and scripts
• Dynamic site configurations
• Search configurations
• Stored queries
• Regenerating or targeted retagging of content
• Alerting authors of new or changed values
• Batch changes and conduct change management
  process in regular cycles
• Analyze change impacts and plan steps

14
Vertical and horizontal metadata

• Horizontal Metadata
• For web sites - 5 Common Look and Feel metatags
  without tools (2000-2006)
• GC enterprise WCMS suite Interwoven, Verity,
  Oracle Portal - provides opportunity for rich
  metadata set for both administering and
  delivering content and information services
• CMS Metadata Element Set and MAP being developed
• RMMES and RM-MAP recently approved
• RDIMS and IM Business Rules initiative
• GC enterprise vision is that we can share core
  metadata
• Delivery Interoperability and government-wide
  provision of information service offerings and
  retrieval, personalization, faceted views, and
  client segmentation
• Administration shared IM business processes,
  leverage development and best practices, and
  GC-wide content reuse and information portability

15
Vertical and horizontal metadata
Without shared metadata
With shared metadata

• Multiple departmental metadata indexes brokered
  at GC level as one virtual index
• Shared business processes, workflows and
  interfaces possible
• Shared cost of development
• Shared development of training materials
• All efforts build on core functionality to get
  the most from investment in solution
• Leveraging content for GC-wide reuse
• GC-wide service offerings

• Multiple departmental metadata indexes
• Mapping and equivalencies at GC level difficult
  if not impossible
• No shared processes or interfaces
• Labour-intensive and costly development with no
  investment reusability
• Difficult to go beyond the basics
• Ongoing web site silos

16
Vertical and horizontal metadata

• Vertical Metadata
• ECM concept covers all aspects of IM including
  documents management and authoring, publishing
  and web delivery, search, records keeping, ATIP,
  correspondence management, accountability, etc
• Harmonized metadata for DM, CM, RM even though we
  dont yet have linked/integrated technologies or
  processes
• Departments will be able to use the CM, RM and DM
  functional elements as core building blocks to
  integrate IM vertically
• GC enterprise CMS-MAP moves us beyond
  department-specific or application-specific silos
  and becomes a key enabler for decentralized
  sharing
• Connects authors, content and presentation
  objects, technical systems and interfaces,
  business processes, workflows and system events,
  content repositories, publishers, web sites,
  search indexes, dynamic web pages, and end-users
  in the IM life-cycle

17
Current and Future Challenges

• We need to start demonstrating the value-added
  services of metadata, or risk losing business
  support and momentum
• Metadata governance, quality assurance and change
  management processes are key
• The metadata will only be value-added if it is of
  high quality and usefulness
• Change management processes will be key to
  success as changing metadata in the MAP is the
  easy part
• The change management processes in a functioning
  CMS can be complex, and departments should plan
  for adequate time, tracking, process cycles and
  governance issues

18
Current and Future Challenges

• We are in a race to finalize the CMS-MAP before
  critical mass of departmental WCMS
  implementations and tagged files are in place
• The MAP can be implemented by any department
  regardless of WCMS status, as it can still be
  useful for interoperability and standardization
• Departments may need to add extensions to the MAP
  to incorporate specific additional requirements
• Perfect the MAP as much as possible prior to
  giving it to implementers, and before content is
  tagged
• MAP design currently abstract and on paper,
  possible difficulties to encode and implement in
  the enabling technology

19
Current and Future Challenges

• Content objects for repurposing and reuse
• Will require metadata at the content object
  level, and increase issues of metadata management
• Authoring templates
• Metadata registry for namespace, elements, MAPs
  and schemes
• XML repositories
• Combination of decentralized and centralized
  metadata authoring

20
Contact Information

• Lynn Herbert
• Project Manager
• Enterprise Information Management Directorate
• Innovation, Information and Technology Branch
• Service Canada
• lynn.herbert_at_servicecanada.gc.ca
• 994-0688
• www.servicecanada.gc.ca