Thursday, November 8, 2012

Good Relations and Schema.org


Today, we are pleased to announce that we are integrating e-commerce schemas from the  GoodRelations project into Schema.org. The addition of these widely adopted schemas into schema.org will make it more easy for Web publishers to express structured data about products, offers, companies, stores and related facts.


Schema.org was designed as a platform where the Web community can come together and share structured schemas that improve the ability for search engines to understand the content of Web pages. Our collaboration with GoodRelations exemplifies this: GoodRelations provides a rich, well known and widely used terminology for e-commerce data sharing. By integrating GoodRelations into schema.org, we make it easier for publishers to adopt, and also combine such vocabulary. Just as with IPTC rNews previously, and many other collaborations, our approach has been to bring together existing work in a way that hides multi-schema complexity behind a common datamodel. 

Effective immediately, the GoodRelations vocabulary (http://purl.org/goodrelations/) is directly available from within the schema.org site for use with both HTML5 Microdata and RDFa. Webmasters of e-commerce sites can use all GoodRelations types and properties directly from the schema.org namespace to expose more granular information for search engines and other clients, including delivery charges, quantity discounts, and product features. Enumerated lists of values remain managed at GoodRelations URLs, following our general approach for referencing 'external enumerations'.

Integrating these schemas has involved making a few decisions, and we welcome all feedback on the approach taken here. In order to have consistent naming conventions between schema.org and GoodRelations, some terms were given new names for use within schema.org. There are also a few cases where existing schema.org vocabulary differed in terminology or 'level of detail' with GoodRelations. We will continue to improve our documentation, examples and FAQ to make clearer the new expressivity that these additions bring. But we wanted to share this progress as early as possible, since it provides an important step forward for structured data and e-commerce on the Web.

Good Relations has been developed and maintained by Martin Hepp since 2002 and continues as an ongoing project. We look forward to seeing it reach new audiences via schema.org.

R.V. Guha

Google





7 comments:

  1. Instead of having webmasters having to change their HTML coding on their website to apply microdata tags, would it not be easier and more efficient to gather the semantic microdata utilizing some type of third party application? This is especially true with user generated content where an end user will upload to a server a Blog, photo, review, etc. but not have any ability to apply microdata tags. The vast bulk of "new" information on the Internet today is Web 2.0 driven, surely there needs to be a way to allow these content producers a means to join the semantic web. The application could then be make its data available to be "harvested" by search engines with microdata tags attached. I threw together a simple BETA application that demos this principle, you can view it and enter sample data at:

    http://www.melbabooks.com/testnation/introduction.html

    The search harvester is at:

    http://www.melbabooks.com/testnation/search.html

    Let me know if you think this is workable!

    ReplyDelete
  2. Thanks for your comment. The main idea behind schema.org, however, is to allow operators of Web site to preserve the meta-data that is already available in their databases so that search engines and other consumers of Web content do not have to apply guesswork for extracting structured data.

    It is far from trivial to properly reconstruct the original data structure from HTML, so the best known approach as of today is to add meta-data markup to the HTML templates of dynamic and static Web applications.

    Form-based generators can help site owners to author respective markup, and I understand that your tool is in line with such approaches. There are others, namely

    http://www.ebusiness-unibw.org/tools/grsnippetgen/

    and

    http://schema-creator.org/

    However, this approach works only for static content (and as a teaching aid for understanding the principle). The far more interesting case, however, are dynamic Web sites that generate thousands of pages for individual products, events, etc. from a database.

    In those cases, there is no way around marking up the HTML template in the traditional way.

    Best wishes

    Martin Hepp

    http://www.heppresearch.com/

    ReplyDelete
    Replies
    1. I agree and am looking into both Drupal and Wordpress. Drupal already has support for many popular vocabularies, out of the box and an extension that works with schema.org. There isn't a specific gr extension/plugin but there is a tool that lets you import various vocabularies. I didn't see on the schema.org site exactly how, specifically, to incorporate GR with schema.org. Do we include both in the RDF mappings?
      Thanks,
      Bruce

      Delete
  3. You can see implementation of microdata with this site of wine events in France
    . We use this site for publish all events (schema Event) about wine, and you can be informed(schema Article) of wine.

    ReplyDelete
  4. Has there been any study done/published that discusses schema (or ontology) overlap between schema.org and goodrelations? It will be interesting usecase to study how one can search across datasets marked up by using either of the two ontologies. I am currently considering starting an experiment for doing this. Not yet sure where to start and what data sources will be appropriate.

    If anyone else is interested in this, please get in touch! Email: ravish.bhagdev@gmail.com

    ReplyDelete
  5. Would there be a difference from Good Relations Validator compared with the others like Yahoo and Bing Validator?

    ReplyDelete
  6. I agree and am looking into both Drupal and Wordpress. Drupal already has support for many popular vocabularies, out of the box and an extension that works with schema.org. There isn't a specific gr extension/plugin but there is a tool that lets you import various vocabularies. I didn't see on the schema.org site exactly how, specifically, to incorporate GR with schema.org. Do we include both in the RDF mappings?
    Thanks,
    Bruce

    ReplyDelete

Note: Only a member of this blog may post a comment.