FRBR and Topic Maps in Designing a Semantic Digital Library

Sam Oh [Professor in the Department of Library and Information Science at Sungkyunkwan University, Seoul Korea] sought to improve the digital catalog of the National Library of Korea. To do this, he sought to apply two models of library cataloging to the corpus: the FRBR model and Topic Maps.

There are two purposes for cataloging: (a) Make a book easy to find and (b) Display books in semantically meaningful groups. Online catalogues are still wrestling with the second purpose. This is Professor Oh’s motivation for applying two cataloging concepts to the National Library of Korea. Both of these concepts rely heavily on associative and relational meta-data to add context to a book record.

Professor Oh produced a new set of records per the FRBR specs by creating an algorithm that harvested the MARC records  and converted them into FRBR records. FRBR records are meant to add a relational aspect to library records, allowing users to take advantage of the associations among entries. As a side note, Professor Oh sees the OCLC’s implementation of the FRBR model for fictional books (FictionFinder) as an exemplary for the National Library of Korea.

Professor Oh admitted that his algorithm only drew associations wherever it could glean it from the MARC records. The end-result was a set of sparse FRBR records.

Prof. Oh then plans to further augment these records by constructing Topic Maps for them.  Topic Maps are composed of  two layers: an information layer and a knowledge layer. The information layer is a collection of index terms much like the terms in the index of a book. The knowledge layer includes topics and associations (akin to the “See Also” section of an index).

The next step for the project appears to be a user study of these new records as a digital system.