University Photographs with embeded metadata

Title

University Photographs with embedded metadata

Description
The collection consists of ~45,000 TIFF images with embedded descriptive metadata. The metadata is consistent, but is (mostly) just a list of keywords. These keywords need to be mapped to the proper Dublin Core metadata fields and, in many cases, the keywords needs to be mapped to different values, like a last name, to a full name, or a 2 letter building code, to the full building name. For example, “ec” needs to be mapped to “Events Center”

Licensing
These photos are all rights reserved and can be used only for this mashup. A set of photos with the same or similar embedded metadata will be provided at a latter time..

Owner
Binghamton University.

Dataset Location
It is on a USB key. Link to current shell script (complete with typos in the comments): http://codabox.org/113/

Dataset Champion
Edward M Corrado

Issues brainstorm

• The first thing that needs to be done is extracting the metadata
• Currently a shell script calls exiftool and a bunch of sed  commands is being called to extract and manipulate the embeded metadata into Dublin Core-(like) metadata
• A configuration file of some sort that can be edited by a cataloging librarian (and not someone who knows sed, would be much more maintainable and extensible.3 things are needed: 1) the keyword embeded in the metdatadata, 2) What the keyword should become, and 3) What dublin core metadata field it should go into.

List of Issues

• Current method is not sustainable
• Need something a non-programer / non-system admin can edit that will create the mappings (see above)

Sample exiftool output:

ecorrado@ecorrado:~/exlibris/rosetta/university_photo/cumulus$ exiftool -X -IPTC:DateCreated -XMP-dc:Description -XMP-dc:Creator -XMP-dc:Subject -XMP-dc:Format -System:Filename *.tif