The Reality is: Most Connections are Proximate
What does it mean to interoperate information on the Web? With linked data and other structured data now in abundance, why don’t we see more information effectively combined? Why express your information as linked data if no one is going to use it?
Interoperability comes down to the nature of things and how we describe those things or quite similar things from different sources. This was the major thrust of my recentkeynote presentation to theDublin Core annual conference. Inthat talk I described two aspects of the semantic “gap”:
- One aspect is the need for vetted reference sources that provide the entities and concepts for aligning disparate content sources on the Web, and
- A second aspect is the need for accurate mapping predicates that can represent the often approximate matches and overlaps of this heterogeneous content.
I’ll discuss the first “gap” in a later post. What we’ll discuss here is the fact that most relationships between putatively same things on the Web are rarely exact, and are most often approximate in nature.
“It Ain’t the Label, Mabel”
The use of labels for matching or descriptive purposes was the accepted practice in early libraries and library science. However, with the move to electronic records and machine bases for matching, appreciation for ambiguities and semantics have come to the fore. Labels are no longer an adequate — let alone a sufficient — basis for matching references.
- The Strongest Link: Libraries and Linked Data (dlib.org)
- Linked Data Needs To Live Up To The Hype, And Soon – semanticweb.com (semanticweb.com)
- The Semantic Web, Linked Data, and Knowledge Organisation (digitalassetmanagement.org.uk)