| OCR Text |
Show Neatrour & Myntti of the University of Utah Marriott Library present…… METADATA PROBLEMS • • • • Size of collections -Close to 300 collections, over 2 million items. Age of collections - started digitizing in 2000 Inconsistent training Variety of collections means different conventions used for collections from: • Internal • External • Campus partners AREAS OF ASSESSMENT VALUES• Is the data consistent with authority sources? • Are we conformant with the MWDL Dublin Core application profile? • Does the data reflect internal best practices? DUBLIN CORE MAPPING - Is it consistent? MISSING DATA- Does each record have required information (eg dates and subjects)? MAN CURRENT PROCESS AGED Or, How To Clean up Data That is Up to No Good METADATA MANAGEMENT TOOLS • CHECKING VALUES- Export collection data as CSV, import into Excel, filter to find missing values • DUBLIN CORE MAPPING - DPLA Aggregation Tools from NCDHC, Modified by MWDL • RECONCILIATION - Vendor based or within OpenRefine • SOLR Index - CONTENTDM collections exported to SOLR, easy to bulk download and query • Student auditing, utilizing tools • Multipart spreadsheet to capture issues across collections • Librarian review • Select collections for clean up work • Prioritize work on collections with missing values and MWDL Dublin Core application profile issues Clean Up Techniques • Mass edit items when possible Individually edit items Script against CONTENTdm desc.all file • • Future Directions • • Upcoming systems migration makes clean-up more urgent Work on Western Name Authority File provides more opportunities for authority control |