The Problem with Records

In the past several weeks, I’ve been fortunate to have attended several one day conferences. In this post, I want to highlight a presentation by Diane Hillman called “Moving to the Open World: It’s not as scary as it sounds”. This presentation was on how linked open data can work for library data in terms of making our data accessible and discoverable by others. This was a fantastic presentation and one of Diane’s best. Out of all the information in the presentation, I want to highlight an important theme that not only Diane spoke about but that has come up before in reading about library data. This is that using xml to describe items is not very different from MARC records. To put it differently, with xml, we create records or documents that are hard to share, expose and provide relationships for the pieces in the xml record. On slide three of Diane’s presentation, she explains that xml assumes a closed world and the xml record is the equivalent to a metadata record or MARC record. What does she mean? For library data and when xml is implemented, the large majority are implemented according to what is called a schema and written in XSD or extensible schema definition. A XSD outlines the requirements that an XML needs to follow to conform to that XSD. Let’s take the analogy of languages and dictionaries. To write effectively and clearly in German, it’s good to know how to spell words and basic grammar. This is syntax and semantics, which is outlined in an xsd. To return to the world of libraries, to implement the metadata standard of MODS or Metadata Object Description Schema is to create an xml file that follows the requirements outlined in the MODS xsd. And we are familiar with this type of analogy. To create a record in my ILS, I have to follow the MARC and RDA standards. Just like a MARC record, one in XML has text in it. For example, a Dublin Core record has the elements in the angle brackets and textual information between the opening and closing angle brackets. We can move these XML records around through harvests or data migration. We can transform them into different XML files according to different metadata standards. We can display the information in the XML. This is no different from what we do with MARC records. And this is the closed system. We need to break out of this closed system. Of course, RDA attempts to do that with its focus on relationships. Another way is to use linked data. To varying degrees of success this can be implemented in XML. There is the MODS RDF ontology. Though there are a few linked open data projects mostly by large research universities like Stanford and Cornell, I am not aware of any institution using MODS RDF ontology. However, MODS RDF along with BIBFRAME express the need to make data open and shareable on the web. I would definitely recommend Diane’s presentation.

Advertisements

Leave a comment

Filed under cataloging, Linked data

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s