Subject authority control, a once taken-for-granted principle of resource discovery, is a somewhat neglected topic that the recent Biennial International UDC Seminar in Lisbon raised to prominence at a two-day meeting on the 29th-30th October 2015.
One hundred delegates from twenty six countries, many of them familiar faces at ISKO events, attended the Seminar. The programme comprised twenty presentations and six posters. Familiar names such as Marcia Zeng, Dagobert Soergel and Douglas Tudhope contributed to the presentations.
Barbara Tillett’s address reflected a lifetime’s work at the Library of Congress. Particularly topical in her discussion were mapping issues. She referred to the MACS project (Multilingual Access to Subjects) and though one-to-one mapping could never be systematically achieved, she observed, an interesting fact was that the range of terms assigned from the multiple language systems provided useful suggestions for users to explore concepts.
The editor of Dewey Decimal Classification, Rebecca Green, and the editor of Universal Decimal Classification Aida Slavic, spoke on the many applications of powerful notational classification schemas that have been employed world-wide for over a century and their potential for applications of linked data. Notation transcends language barriers and simplifies some information challenges posed by linked data it was argued, however it requires leverage to better integrate it into the current information environment.
Of particular interest to thesauri users is the work of Andreas Ledl from the University of Basel, Switzerland and his ambitious register of thesauri, ontologies and classifications.
The work of Ulf Schoneberg and Wolfram Sperber from the FIZ Karlsruhe – Leibniz-Insitut fur Informationinfrastruktur which adapted machine learning methods to meet the specific requirements of mathematical information appeared to have exciting implications for automatic indexing of variables
The potential of automatic classification software was explored by Attila Piros and a method of evaluating automatic subject indexing and classification was presented by Koralijka Golub, Joacim Hansson, Dagobert Soergel and Douglas Tudhope.
We presented results of our research into an application of the Universal Decimal Classification scheme for addressing the relationship between the DDI fields of <subject> and <keyword>. In a recent pilot study at the UK Data Archive the UDC schema proved economic to use for the classification of studies. The potential of classification schemes to work in a linked data environment underpins this work.
A unifying theme of the meeting was that at a time when subject cataloguers and indexers are under continuous pressure to justify the value of their work, success stories in subject retrieval can be linked to both the use of classification and the availability of subject control tools. The quality of these tools, as we are currently experiencing in our thesauri revision work, relies heavily on the expert knowledge of subject cataloguers and indexers.