My colleague Rob has spent a couple of weeks putting together some his thoughts on how we can structure bibliographic data semantically using RDF. The paper was written as a submission for the Linked Data on the Web 2008 workshop @ WWW2008. For me what’s great about this paper is that it provides an insight into the excellent work Rob has been doing in trying to find relationships in MARC data, his work allowed us to do some pretty amazing things with Bibliographic Data last year and we are now, as an organisation, building upon that in some of our commercial applications. At it’s, core though, the issues that the paper covers are not specific to just MARC Data, in order to create linked data we need to start using Links and not just Literals. Whilst Rob was writing this paper, I was busy trying to take some data we had been given and import it into our platform, with Rob’s help I found that by applying the techniques that he describes in the paper and thinking about the granularity of the data itself and the linkages I could create by generating URI’s to represent identities an concepts rather than simply storing those elements of the data as Literal’s – I found that I could exploit relationships in the data that I would not have been able to otherwise.

If anyone has any thoughts or feedback on the paper please let us know.