Survey Contents

1. Design and Planning of Digital Projects

2. Element Set Standards (data structure decision)

3. Data Contents in a Record (data content decision)

4. Authority Files and Controlled Vocabularies (data value decision)

5. Metadata Encoding (data format/ technical interchange decision)

6. Open Questions

Summary 1

Summary 2

Appendix A: Workflow Chart

Appendix B: The questionnaire


Open Questions Summary 1:
Significant issues crossing all questions

Suggestions of major metadata issues from the comments of Open Questions

 
Standardization and interoperability
-All levels should consider using standards: structures, formats, tools, and products.
-Levels of interoperability should be not only syndetic but also semantic (implying not only the data elements and fields be crosswalks but also that the values be correctly converted and exchanged).
-Sharable data should be produced and provided, including descriptive data, subject vocabularies, and even file-naming conventions.
-Metadata for Web archiving and publishers� metadata should be included.
Extensibility
-Decisions should be made whether to create extension elements or separate schemas and only extract the useful elements.
Multilingualism
-It is important to consider correct character sets for encoding non-Roman languages.
Quality vs. efficiency
-Quality of metadata, especially in the non-MARC format or input by nonprofessionals, became a clearer issue than was previously realized.
-Metadata creation is a costly process. Metadata production consumes enormous amounts of time.
-It is still not clear how to calculate the hidden costs associated with different metadata decisions.
-Metadata architecture should be studied to explore harvesting models and query models so that metadata can be shared and used efficiently and automatically.
-Among the more specific comments, respondents suggested:
~the need to explore how to capture metadata in the most efficient way, e.g., generate values automatically, reduce the number of mandatory fields for metadata creators, harvest from other repositories
~the need to explore ways to introduce user-generated metadata, e.g., tagging, reviews, and how best to incorporate it with traditional metadata
~the need to get search engines to handle metadata so that users get the greatest benefit
Staffing and Training
-Good metadata creators are in high demand. Training is critical to not only nonlibrary professionals and noncatalogers, but also the catalogers who have been trained only in more traditional conventions.
More open and flexible choices
-Strongly encourage exploring non-MARC format, as emerging (and in some cases, established) standards for the creation of document structures and metadata provide greater flexibility and better integration with mainstream software applications such as enterprise-scale databases.
-Suggest discovering "how to export and share metadata from a digital project into an aggregated environment - either our own aggregation, or as part of a larger community. Beyond OAI/Dublin Core!"
 See also Summary 2.