Enhancing the Quality of Metadata: Modular Approach to Digital Resource Lifecycle Management

1
Enhancing the Quality of Metadata Modular
Approach to Digital Resource Lifecycle Management

• Daniel Gelaw Alemneh Mark E. Phillips
• IST, Archiving-2007 Conference
• May 23, 2007, Arlington Virginia

2
University of North Texas (UNT) Libraries Digital
Initiatives

• Collaborative Initiatives
• CyberCemetery
• GPO
• NARA Affiliated Archive
• Texas Register Archive
• Secretary of States Office
• Texas Laws and Resolutions Archive
• Secretary of States Office
• The Portal to Texas History
• 45 Libraries Museums
• Web-at-Risk Project
• California Digital Library
• New York University

3
University of North Texas (UNT) Libraries Digital
Initiatives

• Library Digital Collections
• Congressional Research Service Archive
• 9500 CRS Reports
• Portal to Texas History
• 20,000 records
• World War Poster Collection
• 500 WWI and WWII Posters
• Advisory Commission on Intergovernmental
  Relations
• 408 reports 47,874 pages
• Federal Communications Commission (FCC) Record
• 136 issues 43,115 pages (6 of 21 volumes
  completed)
• GovDocs A to Z digitization project
• 186 scanned 500 in queue
• Jean-Baptiste Lully Collection
• 27 scores 10,000 pages

4
Metadata Environment

• Metadata-based digital resource management
  activities
• UNT Libraries metadata locally qualified Dublin
  Core based descriptive metadata.
• Detailed technical and preservation metadata
  elements
• Web based metadata creation and editing
• Interoperability
• Metadata Crosswalks
• Mods
• Marc
• oai_dc
• PREMIS

5
Metadata Quality

• The two aspects of digital library data quality
• The quality of the data in the objects
  themselves
• The quality of the metadata associated with the
  objects
• Poor metadata quality
• Ambiguities
• Poor recall
• Poor precision
• Inconsistency of search results

6
Metadata Quality

• Most Common errors
• Incorrect Data
• Letter transposition
• Letter omission
• Letter insertion
• Letter substitution or misstrokes
• Missing Data
• Elements and values not present at all (null)
• Insufficient or incomplete data
• Ambiguous Data
• Confusing or inconsistent data e.g. multiple
  spellings, multiple possible meanings, mixed
  cases, initials, etc.

7
Factors Influencing Metadata Quality

• Local Requirements
• Objects Heterogeneity
• What type of objects will the repository
  contain?
• Granularity
• How will they be described?
• Functionality
• What functionality is required?
• How will it be interfaced?

8
Factors Influencing Metadata Quality

• Collaborative Requirements
• Diversity of Users
• How best diverse information-seeking behaviors
  can be met?
• Interoperability
• Will metadata be meaningful within aggregations
  of various kinds?
• What is required for interoperability?
  (Structure, semantics, syntax)
• Digital rights issues
• Will access restrictions be imposed?
• Are requirements formal or informal?
• Are there other access and associated digital
  rights issues?

9
Factors Influencing Metadata Quality

• Training Issues
• Necessary expertise to create and manage
  rigorous metadata
• Metadata quality can be determined to a great
  extent by
• knowledge of the source, and
• knowledge of the methodology used to create the
  statement
• Cost
• Rigorous metadata is resource intensive and too
  costly

10
UNT Metadata Quality Assurance Mechanisms Tools

• The two main stages of metadata qualities
  assurances
• Pre-injust
• 1. Metadata Creation tools (Templates)
• Post-injust
• 2. Metadata Analysis tools (Web-based tools)

11
Quality Assurance Mechanisms and Tools Templates

• Metadata Creation Tools (Templates)
• Validates Mandatory elements
• Metadata Template Creator
• Template Reader
• Controlled vocabularies (UNTLBS)

12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
UNT Metadata Quality Assurance Mechanisms Tools

• 2. Metadata Analysis Tools
• NULL Values
• List/Browse All Values (by each qualifiers and
  elements)
• List Authorities Values
• Graphical reports and other fun stuff
• Clickable Maps by Institution and Collection
• Word Clouds by elements
• Records added overtime and other graphical
  reports

26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
Summary

• Determine level of quality required
• Partners may have much in common, but they have
  diverse and sometimes conflicting metadata
  requirements.
• Determine nature of gap and how to close it
• effectiveness, efficiency, practicability,
  scalability
• Machine verses human error handling
• How much of the process can be automated?
• Human review of results is still essential (e.g.
  highlighted items)
• Compromise
• One size does not fit all!
• Prioritize
• Resources very unlikely to be available to meet
  all requirements
• Test the workflow
• Test, retest, and evaluate the quality cycle
  continuously

41
(No Transcript)
42
Questions?