Matthew Beacon said:
>Transcription (of a sort) would continue to be how we often get the
>description, but it would be machine transcription from digital
>source files (e.g metadata from headers in tiff files, metadata from
>XML documents already marked up by creator/publisher.)
Our experience of metadata as received from clients is that
correspondence with the resource is sometimes very slim. I'm all for
electronic retrieval of data, but it should be from the resource (in
our case often a PDF file), not from metadata created by who knows
whom. Even standard numbers may be wrong. Abbreviations and
misspellings (judgement/judgment) abound. Homonymns are not unknown.
Strange entities may be considered "Author". The title may lack
beginning words. Metadata is not always created by electronic
derivation from the resource.
Our experience with character recognition is not very hopeful.
Electronic recovery of data from scanned material is very problematic.
__ __ J. McRee (Mac) Elrod ([EMAIL PROTECTED])
{__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/
___} |__ \__________________________________________________________
