Loading ...
Sorry, an error occurred while loading the content.

Relation between REST and Dublin Core/XMP

Expand Messages
  • javatech
    This is a newbie open-ended question. So if this is too generic I will read up more and come back. I am trying to understand how REST and metadata initiatives
    Message 1 of 3 , Aug 26, 2010
    • 0 Attachment
      This is a newbie open-ended question. So if this is too generic I will read up more and come back.

      I am trying to understand how REST and metadata initiatives are related to each other. Why do you need Dublin Core/XMP etc. ? Are these microformats ?
    • Brian Sletten
      ... It is a big, open question, but I ll try to push you in the right direction. ... On the surface, there is no direct connection except the Web architecture.
      Message 2 of 3 , Aug 26, 2010
      • 0 Attachment
        > This is a newbie open-ended question. So if this is too generic I will read up more and come back.
        It is a big, open question, but I'll try to push you in the right direction.

        > I am trying to understand how REST and metadata initiatives are related to each other. Why do you need Dublin Core/XMP etc. ? Are these microformats ?

        On the surface, there is no direct connection except the Web architecture. REST is an architectural style for managing information resources. It has very little standard metadata associated with it other than that it inherits from HTTP.

        Dublin Core is a framework for describing publication metadata. It was produced by a bunch of librarians through the OCLC in Dublin, OH. It originally started around the Warwick Framework but was recast as the poster child for RDF along the way. Dublin Core is mostly used to describe authorship, subject designation, publication dates, etc. It is actually a more complicated framework that supports interoperability across metadata profiles, but for your purposes here, it is an RDF vocabulary for describing resources with standard metadata terms (dc:title, dc:subject, dc:creator, etc.)

        It would be used either directly as RDF:

        http://bosatsu.net/index.html http://purl.org/dc/terms/creator http://purl.org/net/bsletten

        This is a simple fact or "triple" connecting a document to an author, indicated by a 303 non-network-addressable resource through the Dublin Core creator relationship). RDF statements follow a subject - predicate - value relationship but can have many different serializations. In this case, both the subject and the relationship are global and resolvable:

        http://purl.org/dc/terms/creator

        This can resolve both human-readable and machine-processable versions of the relationship. The data model allows you to use relationships from other vocabularies so it makes it very easy to accumulate data from the Web. People are now starting to weave RDF into XHTML, HTML, SVG, ODF, etc., generating it on the fly, exposing it natively as part of the Linked Data Project (http://linkedata.org).

        There are technologies that build on RDF such as SKOS and OWL to allow you to organize the terms and resources in new and interesting ways. You can then start to do certain types of inference over the data organized this way. One of the exciting parts is that you can organize other people's data the way you want to see it relatively easily.

        RDF and microformats serve similar goals (to describe documents and resources) but they have much different scopes. RDF has a data model associated with it and is largely intended to support global references and relationships. Microformats are intended to be simple, developer-friendly ways of encoding certain domains (events, people, reviews, organizations, etc.)

        The good news is that it is easy to convert Microformats into a form that can be used with RDF so it is all good metadata.

        XMP is based on an older version of RDF and was intended as a way of allowing Adobe's various partners to contribute tools in a document-processing framework and allowing them all to annotate a document, image, etc. with metadata (camera information, filters applied, etc.) It isn't super-wildly used but I think the adoption of RDFa by ODF is going to help spur interest here again.

        The excellent "RESTful Web Services Cookbook" and "REST in Practice" books touch upon the relationship between REST and Semantic Web technologies like RDF, but I am taking a much deeper dive in a book I am writing for Addison-Wesley called "Resource-Oriented Architectures : Building Webs of Data".
      • Eric J. Bowman
        Great post, looking forward to the book. BTW, the correct link is http://linkeddata.org/ . -Eric
        Message 3 of 3 , Dec 11, 2010
        • 0 Attachment
          Great post, looking forward to the book.

          BTW, the correct link is http://linkeddata.org/ .

          -Eric

          Brian Sletten wrote:
          >
          > > This is a newbie open-ended question. So if this is too generic I
          > > will read up more and come back.
          > It is a big, open question, but I'll try to push you in the right
          > direction.
          >
          > > I am trying to understand how REST and metadata initiatives are
          > > related to each other. Why do you need Dublin Core/XMP etc. ? Are
          > > these microformats ?
          >
          > On the surface, there is no direct connection except the Web
          > architecture. REST is an architectural style for managing information
          > resources. It has very little standard metadata associated with it
          > other than that it inherits from HTTP.
          >
          > Dublin Core is a framework for describing publication metadata. It
          > was produced by a bunch of librarians through the OCLC in Dublin, OH.
          > It originally started around the Warwick Framework but was recast as
          > the poster child for RDF along the way. Dublin Core is mostly used to
          > describe authorship, subject designation, publication dates, etc. It
          > is actually a more complicated framework that supports
          > interoperability across metadata profiles, but for your purposes
          > here, it is an RDF vocabulary for describing resources with standard
          > metadata terms (dc:title, dc:subject, dc:creator, etc.)
          >
          > It would be used either directly as RDF:
          >
          > http://bosatsu.net/index.html http://purl.org/dc/terms/creator
          > http://purl.org/net/bsletten
          >
          > This is a simple fact or "triple" connecting a document to an
          > author, indicated by a 303 non-network-addressable resource through
          > the Dublin Core creator relationship). RDF statements follow a
          > subject - predicate - value relationship but can have many different
          > serializations. In this case, both the subject and the relationship
          > are global and resolvable:
          >
          > http://purl.org/dc/terms/creator
          >
          > This can resolve both human-readable and machine-processable versions
          > of the relationship. The data model allows you to use relationships
          > from other vocabularies so it makes it very easy to accumulate data
          > from the Web. People are now starting to weave RDF into XHTML, HTML,
          > SVG, ODF, etc., generating it on the fly, exposing it natively as
          > part of the Linked Data Project (http://linkedata.org).
          >
          > There are technologies that build on RDF such as SKOS and OWL to
          > allow you to organize the terms and resources in new and interesting
          > ways. You can then start to do certain types of inference over the
          > data organized this way. One of the exciting parts is that you can
          > organize other people's data the way you want to see it relatively
          > easily.
          >
          > RDF and microformats serve similar goals (to describe documents and
          > resources) but they have much different scopes. RDF has a data model
          > associated with it and is largely intended to support global
          > references and relationships. Microformats are intended to be simple,
          > developer-friendly ways of encoding certain domains (events, people,
          > reviews, organizations, etc.)
          >
          > The good news is that it is easy to convert Microformats into a form
          > that can be used with RDF so it is all good metadata.
          >
          > XMP is based on an older version of RDF and was intended as a way of
          > allowing Adobe's various partners to contribute tools in a
          > document-processing framework and allowing them all to annotate a
          > document, image, etc. with metadata (camera information, filters
          > applied, etc.) It isn't super-wildly used but I think the adoption of
          > RDFa by ODF is going to help spur interest here again.
          >
          > The excellent "RESTful Web Services Cookbook" and "REST in Practice"
          > books touch upon the relationship between REST and Semantic Web
          > technologies like RDF, but I am taking a much deeper dive in a book I
          > am writing for Addison-Wesley called "Resource-Oriented
          > Architectures : Building Webs of Data".
        Your message has been successfully submitted and would be delivered to recipients shortly.