Workshop on Application of Technology and Classification Tools for Digital Collection Development and Network-Based Search
ACM DL '98, Pittsburgh, June 27, 1998
Move forward the initiatives toward interactive networked access to reference tools
e.g., thesauri, classification systems, gazetteers, taxonomies, ontologies, language translation, dictionaries, etc.
Focus on current and planned developments.
Data model to support interactive use.
Data elements and relationships to convey content? Elements (ANSI/ISO or something higher level)
Computer understanding of system notations. Are unique ID for terms/concepts needed? (URI/URN, DOI)
Representational structures (XML)?
Representation of purpose
What user/use interaction requirements - what conversations between client and server?
How to access and navigate?
What are most important features?
Quality standards (CIQM labeling). Breadth, depth, coverage.
Thesaurus-level metadata & thesaurus registries
Metadata representation of tools for registries - "collection level"
Role of metadata to enable interoperability
Role of tool registries
Infrastructure and mechanisms of registries
Business Intellectual Property Models
Types of agreements and partnerships, kinds of business models. What happens when put these in online environment? Multiple tools, multiple environments. Fringes of domain.
What kinds of relationships are needed between owners and users?
Rights management issues and control?
Appropriate access and use constraint models?
Transaction and payment models?
Data Model Group Session
How generic do we need to be beyond thesauri?
Glossary - Agreement on terms we use; what are the objects we are dealing with?
Types of information objects (reference tools)
Thesauri - hierarchical, synonyms
Dictionaries - Lexicons
Taxonomies (a classification)
Lists of valid terms
Data dictionaries and object hierarchies
A typology such as this will be a necessary for this effort.
Qualities of successful data model
Authority (validating source)
History note (temporal information)
Classification (Facet, Categories)
Representation (e.g., language, character set)
Identification (notation, etc.)
Directed graph typing
The types defined by the NISO, ISO, etc. standards
Terminology services (whole tools - with specific access)
Concept expansion / focusing
Limiting context of use/view
Filtering (ranking, document clustering)
Scenario: supporting interactive use
Record w/ "anchor" e.g. pollution
Want to find related information
Pollution + DDC + LCC+ related concept + related organizations
What you want to do:
identify tools with related information (metadata for resource discovery)
browse, search tools
retrieve terms from tools to reuse terms in follow-on processes
Effective Next Steps
Scenarios of use
More specific models based on scenarios
Glossary of terms
Develop standards for use of tools in network search (directly and indirectly)
Demonstrate an interest in forming a group. Setting the (functional) requirements before we set the standards.
Commitment to work offline, to make information publicly available (accessible)
Thesaurus registry (Interconnect) - broaden information types
Use form as straw man.
Taxonomy of information types, definitions
External knowlwedge sources (Buckland and Plaunt) that's a component of a larger search system. Reference model. Other models? This sort of stuff could be mounted/pointed to at web site.
Forum for review and collaboration.
Planning new products.
Form task forces/committees
Developing interactive thesauri
Interface issues - Interoperability
Registries for Knowledge Organization Schemes (KOS)
White paper report on the scope and requirements of the project
Web site providing links to related projects and sites and adding additional information
Project glossary, authoritative source of terminology, on the Web
Plan next workshop – who what when where why