Special Issue
Government Information Quarterly

Metadata
A Networked Information Strategy to Improve
Access to and Management of Government Information

Draft Articles' Abstracts and Outlines

Special Issue Editor
William E. Moen
<wemoen@unt.edu>
School of Library and Information Sciences
Texas Center for Digital Knowledge
University of North Texas
P.O. Box 311068
Denton, Texas 76203-1068
Voice: 940-565-3563
Fax: 940-565-3101

 

Contributors
Eliot Christian
<echristi@usgs.gov> 
US Geological Survey
802 National Center
Reston VA 20192
Voice: 703-648-7245  
Fax: 703-648-7112 

Anne Craig
<acraig@ilsos.net>
Associate Director, Library Automation & Technology
The Illinois State Library
300 S. 2nd Street
Springfield, IL 62701-1796
Voice: 217-785-5607

Allen Mullen
<amullen@tsl.state.tx.us>
Developer, Networked Services
Texas State Library and Archives Commission
P.O. Box 12927
Austin, TX  78711
Voice: 512-463-5534.
Fax: 512-936-2306

Eileen Quam
<eileen.quam@dnr.state.mn.us>
Information Architect
Minnesota Department of Natural Resources
Management Information Services 
500 Lafayette Road N 
St. Paul, MN  55155 
Voice:
651-297-2341


[Working Title]
Introduction to the Special Issue
William E. Moen

[Note:  An abstract and outline for the introductory essay will be developed after I receive the first complete drafts of the articles by the contributors.  Generally, I'll lay out what metadata is, why it is important, what it is used for, etc.  The introduction will also introduce the articles by the contributors.]


 

[Working title]
A Metadata Initiative for Global Information Discovery
Eliot Christian

 Outline 

1. Context and Rationale
      
The Global Information Society
      
The "Information Discovery" Problem
      
The Internet is not a Library, Yet
      
The Vision: a Global Information Locator Service

2. Concepts and Design
      
Metadata as "Characteristics"
      
Semantic Mapping
      
Locator Records
      
Interoperability for Information Discovery
      
Common Search Concepts
      
Designing for Diversity

3. Practical Implications
      
A Searcher Perspective
      
A Content Provider Perspective
      
A System Engineering Perspective
       
An Information Infrastructure Perspective  

4. Organizational Approach
      
Information Communities
      
Organizing an Information Community
      
Intersections among Information Communities

5. Example Initiatives
      
Bibliographic Information Communities
      
Geographic Information Communities
      
Environmental Information Communities
      
Government Information Communities


[Working Title]
Anne Craig

Abstract

The argument for the necessity of using a controlled vocabulary for facilitating information retrieval on the Web provides the premise for an account of Illinois' customization and implementation of lexicographer Jessica Milstead's subject tree for the Find-It! Illinois Program.  With the vast wealth of Illinois government resources available on the Web, the application of a subject thesaurus was viewed as essential for maximizing access and ensuring relevant search results.  Since the value of information is inherent in its utility, providing the best possible access to information remains the ultimate mission of libraries.  Certainly, in a time characterized by the proliferation of information, abandoning controlled subject access in favor of keyword "hit or miss" seems irresponsible and deceptively simplistic.  To further complicate the issue, the existing cataloging tools such as LCSH and other subject schemes cannot merely be applied to the Internet.  Since the information is retrieved within this virtual medium, the need for the development of new tools and standards in order to improve online browsing capabilities is evident.

Toward that end, an analysis of Jessica Milstead's subject tree for Find-It! Illinois addresses the following topics: 1). rationale for the structure of the subject tree; 2). popular usage vs. established headings; 3). the importance of cross-references; and 4). the capability to be both universal and customizable to facilitate interoperability. The implementation of the Find-It! Illinois subject tree has led to the goal of nationwide expansion. While an exciting prospect, the nationwide adoption of a standard controlled vocabulary for accessing government information on the web will require much work and collaboration with other states.  To conclude, an account is presented of Illinois' work with other states to coordinate and customize the Find-It! Illinois subject tree so that the vision of a "Global Find-It!" may one day be actualized.


[Working Title]
Metadata Initiatives in States
Allen Mullen

Abstract

Metadata has developed as a means of providing access and  managing information resources for Government Information Locator Services (GILS) in several states.  A variety of metadata schemes have developed building on efforts in the federal government, and the library and information science community.  The approaches to metadata include incorporation into the resources, metadata wrappers, and development of repositories to store metadata separate from the resources they describe.  Metadata use is also developing as a strategy for preservation of resources.  Developments in RDF, XML and Z39.50 provide means for integrating resources using various metadata schemes into seamless access across diverse resources.

Outline

I. History of metadata in GILS

Federal GILS records 
State adoption of GILS metadata
Federal role/leadership
Hybrid GILS records - WAGILS and derivatives
Dublin Core
Other GILS metadata models

II. Methods of metadata development

Storage
Embedded in document
Repository 
Metadata wrappers 
Agency generated
"Cataloger" generated
Harvested
Metadata wrappers

III. Uses of metadata across various state models

Public discovery of published information
Records access
Administration
Preservation

IV. GILS integration

Z39.50 
RDF/XML
Z-Tokens

V. Future of metadata use in GILS

Metadata developments and XML
Metadata use


[Working Title]
Foundations Project Needs Assessment and Usability Studies
Eileen Quam

Abstract

A number of studies were conducted during the course of the development of the State of Minnesota Foundations Project's Bridges web site.  This article discusses the purpose, methodology, and results from the preparatory needs assessment study, the Bridges user interface study, and most importantly, the usability study on Dublin Core metadata and controlled vocabulary.

The grant-funded Foundations Project was designed as a multi-agency collaboration aimed at improving public access to environmental and natural resources data and information.  Developing intuitive and easy to use search tools and strategies as the Project's focus, staff and agency participants added Dublin Core metadata to many kinds of electronic information resources including web pages, PDF documents, and geographic data.  Such resources developed advanced search and retrieval techniques that integrate access to this information across state environmental agency web sites.  Based on project research findings and using principles of information architecture, a blueprint has been created that specifies best practices for these web sites.

Each of the studies is discussed, and where applicable, methodology is offered for adaptation by others.  A great deal of usability study research was drawn on in preparation for the studies themselves, and the author offers this information. A number of reports shows further background work that informed the decision-making process for such important choices as metadata schema, thesaurus and thesaurus management software, search engine, and RDF/XML standards.

An overview of controlled vocabulary and thesauri leads to an understanding in the areas of thesaurus choice, thesaurus management software, and use of the vocabulary in Dublin Core metadata as well as in Ultraseek's Content Classification Engine (CCE). CCE is a topic hierarchy creation tool that utilizes rules, including Dublin Core subject tags, to populate the topics. This is relevant to the third study on Dublin Core metadata, which tested keyword searches of web pages to determine the effectiveness of controlled vocabulary in the Dublin Core subject tag. The consistent set of keywords were searched by an outside analyst three times over 2 years, tracking changes in dead-end searches, excessive results, irrelevant results and the like as more metadata was applied to the web pages.

In conclusion, the author discusses the Best Practice Guidelines for Web Metadata, freely available on the Bridges web site. In addition to the reports of the three studies and the background reports, the Best Practice Guidelines include current bibliographies. The bulk of the Best Practice Guidelines is the User Guide to Dublin Core, which offers an element-by-element understanding of the metadata schema, as well as a glossary, visual cheat sheets, and information on TagGen metatag generator.

Outline

1.    Foundations Project
        a. Overview
        b. Purpose
        c. Tenets
        d. Standards chosen, including Dublin Core
        e. Software chosen
        f.  Background documentation to decisions made

2.  Needs Assessment Study
        a.  Purpose
        b.  Methodology
        c.  Conclusions

3.  User Interface Study
        a. Purpose
        b. Methodology
        c.  Conclusions

4.  Vocabulary Control / Thesauri
        a.  Overview
                i.  What is controlled vocabulary
                ii. What is a thesaurus
                iii. What are naming conventions / authority lists
                iv. Why needed
        b.  Legislative indexing vocabulary
                i.  What it is
                ii. Why chosen
        c.  Lexico thesaurus management software
                i.  Managing a changing vocabulary
                ii. Editing capabilities
        d.  Ultraseek's Content Classification Engine
                i.  Description (Yahoo-style topic hierarchy) and purpose
                ii. How it works; rule creation
                iii. Incorporating controlled vocabulary
                    1. Topic creation
                    2.  DC Subject rules

5.  Usability study on Metadata and Controlled Vocabulary
        a.  Purpose
        b.  Methodology
        c.  Results and conclusions

6.  Best Practices for Web Metadata
        a. Overview of MN best practices document
        b.  Description of content with links
                i. Minnesota Metadata Guidelines
                    1. User Guide to Dublin Core
  
                 2. DC glossary
                    3. TagGen information
                    4. Visual cheat sheets
                ii. Bibliographies
                iii. Appendices, including user studies
                iv. Background reports

7.  Wrap-up