Hi all,
I'm currently working on the FuzzyPhoto research project at De Montfort University with a series of partner collections (http://fuzzyphoto.edublogs.org/).
We are trying to do cross collection co-reference identification for photographic records (i.e identifying multiple occurrences of the same image) but it has broader search applications. We're trying to get around the limitations of current keyword based search systems with the aim of making discovering similar items much easier for researchers.
However, what has been really apparent is that while collection APIs are making it much easier to access records and whilst shared schemas are slowly pushing people towards shared record formats, the internal field representations have very little consistency. Person names and dates are the obvious examples, how the information is represented within a date field is effectively random once you start trying to do cross collections searching. I think that a big challenge for cross collection searching is going to be how to really understand the field contents.
David Croft
****************************************************************
website: http://museumscomputergroup.org.uk/
Twitter: http://www.twitter.com/ukmcg
Facebook: http://www.facebook.com/museumscomputergroup
[un]subscribe: http://museumscomputergroup.org.uk/email-list/
****************************************************************
|