1	Designing a Metadata Management Repository Michael Pelikan, Penn State University Nathan Rupp, Cornell University Jeff Young, OCLC
2	Many Digital Library Systems
3	A Couple of Metadata “Pieces” Mapping MARC not used in most digital library systems! Various kinds of maps need to be developed: MARCàDC MARCàTEI Transformations Programming scripts that actually carry out the mapping
4	Metadata Management Design Not just recording mapping & transformation processes Also includes: Decision processes that arrived at the mapping and transformation tools Documentation of all this work Metadata management design results in: Promotion of sharing and reuse of tools Recognition that librarians are users too Improvement in operational activities Reduction in the risk that metadata generation and transformation processes will be lost
5	A New Idea? Importance of resource management in general: Unlimited supplies of resources are expensive! To maximize the value and minimize the cost of resources To anticipate and meet the requirements of resources To ensure that resources are used prudently, efficiently, effectively, and securely Data resource management: data as important organizational asset
6	Benefits of Metadata Management Cultural Enhances communication Involves many in furthering the library’s mission Treats metadata as a shared resource Functional Enables easy location of metadata tools created throughout the library Minimizes the creation of redundant components Increases productivity among metadata and IT practitioners Accelerates development of new metadata applications Improves likelihood that new mappings and transformation tools will be easily integrated into the existing metadata environment.
7	Creating a MARC Metadata Management Design Discussion and consensus Costs Benefits Staff involved Documentation and inventory
8	Metadata Inventory (1) MARC bibliographic metadata Script for extracting MARC metadata Resulting file of MARC extract Project specific MARC mapping XML metadata collection and storage scheme Script transforming MARC file to XML Resulting file of MARC-XML transformation
9	Metadata Inventory (2) Script integrating XML with administrative metadata and OCR into TEI Resulting TEI file DTD for validation Metadata as it is stored in the DL system
10	Next Steps Move from MARC repurposing design to metadata management design Reusable transformation tools Investigate the creation of a metadata management repository Sharing metadata repositories
11	We keep thinking we’re the Jetsons…
12	The Big Idea(s) It’s a Repository = Discoverable Resources. Discovered Tools are used. Use = Practices. Tools & Practices that are “in common” gain value as more folks use them.
13	Not just a bit bucket Some kind of descriptive metadata whether a centralized model or distributed, “standardized” description will make the tools discoverable. If distributed, common descriptive practices can make it easy* to harvest / search the “catalog(s)” *yes, easy.
14
15	But why stop there? How about RUD?* Why? Folks could make informed choices among existing works closely fitting their needs. Less duplication of effort! *Real Usable Documentation
16	How about searchable comments? “Well, we started with Cornell’s marc to DC but then added OCLC’s approach to mapping VRA because we had digitized photographs of statuary with…” resolution for the images (dpi), measurements in two-dimensions for the photos (inches), and in three dimensions for the statues (centimeters, plus mass in kilograms)
17	Next thing you know, Folks might start to use stylesheets from the same gene pool After a few generations, there will come to be a “family resemblance” to these things… Interoperability! Best Practices!
18	So, even though we’ve already got standard methods… We might start to see: Less re-invention, institution-by-institution. Less wear & tear - creative solutions can be shared. Easier interchangeability of parts (as a result of employing a rich, growing, but commonly-employed tools set) Just like the Modern Age!
19	XSLTProc model
20	XSLTProc model
21
22