Timothy W. Cole
Mathematics & Digital Content Access Librarian
University of Illinois at Urbana-Champaign
Kenning Arlitsch
Dean of the Library
Montana State University
Ted Fons
Executive Director, Data Services & WorldCat Quality
OCLC
Janina Sarol
Visiting Research Programmer
University of Illinois at Urbana-Champaign
Transforming Bibliographic Records into Linked Open Data (LOD): As an increasing number of institutions seek to transform metadata from their library catalogs into linked open data (LOD), they encounter a number of issues and practical challenges in integrating LOD services into their workflows and transforming MARC bibliographic records into RDF using emerging semantics standards such as Schema.org. The goal is to enhance discoverability of and interactions with library collections on the Web, but how to get there is not always obvious. Ted Fons, will discuss OCLC’s experimentation with new models for library collection metadata management and exposure, describing OCLC’s goals, its data strategy, and the benefits of LOD and entity-based data models. This presentation will describe the community’s effort to better represent bibliographic materials in Schema.org through the Schema BibExtend effort. It will also include a report on experiences and lessons learned in using some of these OCLC resources and services to help the University of Illinois at Urbana-Champaign (UIUC) transform and share bibliographic and holding metadata records describing its 11 million volume print collection as Schema.org graphs, serialized in JSON-LD and HTML+RDFa. We will also briefly describe early UIUC experimentation to better understand additional ways LOD graphs and services might be leveraged to enhance end-user interactions with library metadata.
Establishing Semantic Identity for Accurate Representation on the Web: Libraries both large and small continually miss the opportunity for accurate and robust representation in semantic web applications such as Google’s Knowledge Card display, and possibly suffer from reduced visitation and use as a result. Machines have trouble understanding the concept of “things” and their relationships without considerable help. Google and other search engines are most successful in gathering accurate information about libraries and their services if those things exist as defined entities in the Semantic Web. Accurate recognition as an entity depends on verification from data sources that Google trusts, including Wikipedia, Freebase, etc. Our research demonstrates that libraries and library concepts tend to be poorly represented on the Web, and we can do a better job of improving the semantic data about our buildings, products and services in those trusted data sources.
Learn how Montana State University* and OCLC Research** have been working to refine machine comprehension of the library using Wikipedia, Google Places, Freebase, and Schema.org markup in RDFa and JSON.
* Jason A. Clark, Scott W.H. Young and Leila Sterman from Montana State University have contributed to this research.
**Jeffrey Mixter, Jean Godby and Jeff Young from OCLC Research have contributed to the intellectual content of this research.
http://catalogdata.library.illinois.edu/
Presentation (Sarol)
Presentation (Arlitsch)