I recently had the opportunity to attend the ISKO UK Conference 2013. The conference theme was ‘Knowledge organization – pushing the boundaries’ with a special focus on how better synergy could be achieved between research and practical implementation. The conference itself played an important role in promoting this synergy, since it brought together delegates from across the different sectors (private, public, academic) whose interests ranged from the purely theoretical to practical applications. Leonard Will and Jutta Lindenthal, two eminent thesaurus consultants, spoke about ISO 25964, the new thesaurus guidelines that we are reviewing in our restructuring work on HASSET and ELSST. It was useful to hear their views, and those of other thesaurus experts such as Stella Dextre Clarke and Vanya Broughton, on principles of thesaurus construction.
The relationship between thesauri and other Knowledge Organisation Systems (KOSs) to ontologies was also explored. Vincenzo Maltese from the University of Trento and Ricardo Eito-Brun from the Universidad Carlos III de Madrid both described how they applied methodologies normally associated with KOSs to the development of ontologies. Maltese has used the resulting ontology for document retrieval while Eito-Brun has applied his ontology to the field of requirements specification (in this case, software specifications).
A major theme of the conference was Linked Data (LD) and its applications. Jeremy Tarling and Matt Shearer from the BBC described how LD are being used at the BBC to link news items across the website. Similarly, Daniel Evans and Ailsa Jenkins from the Science Museum, London, described a LD project at the Science Museum to link web pages on related objects and persons on their website. Meanwhile, Jeremy Bentley from Smartlogic Semaphore Ltd demonstrated an LD-based tool that performs predictive analytics on health care data from different authors and sources. All these applications are underpinned by ontologies that they have created and by RDF technology. The planned conversion of ELSST to SKOS in the CESSDA ELSST project, following the conversion of HASSET to SKOS during the SKOS-HASSET project is intended to introduce ELSST to the LD world. As SKOS emerged from the LIMBER project of 1999-2001, which also gave us ELSST, this neatly brings the two together again.
Other talks looked in detail at some KOS-based techniques and applications, including automatic indexing and topic classification (see for example the talks given by Yi Hong of Harris Corporation – Healthcare Solutions, and Andreas Vlachidis from the University of South Wales), thesaurus mapping (from our esteemed ELSST colleague, Andreas Kempf from GESIS, and Christine Laaboudi from the European Union(EU)), and thesaurus quality checking (Christian Mader from Semantic Web Company). We at the UK Data Archive recently conducted an experiment in automatic indexing during the SKOS-HASSET project, and this is something we are interested in pursuing in future. We have already received requests to use the SKOS version of HASSET in thesaurus mapping, and we expect that there will be a lot of interest in ELSST also once the SKOS version becomes available. Laaboudi described the work that is being carried out on thesaurus alignment and management within the EU. Like us, the EU has several thesauri on different platforms, which they are looking to integrate. They are also collaborating with the United Nations to investigate expanding the EUROVOC thesaurus in terms of language and scope via the UN Unbis thesaurus.
All in all, it was a very interesting few days, which has given us all much food for thought. Abstracts and full papers are available on the conference website.