Semantic Knowledge Organisation Systems
The problems of keyword searching are well known. Significant differences in results stem from trivial variations in search statements. These problems can be alleviated by Knowledge Organisation Systems (KOS), such as classifications, gazetteers, lexical databases, ontologies, taxonomies and thesauri. KOS model the underlying semantic structure of a domain for the purposes of general retrieval. They act as semantic road maps and make possible a common orientation by indexers and future users - whether human or machine. Douglas Tudhope was PI on FACET and leads the research on Semantic Knowledge Organisation Systems.
EPSRC FACET Project
The FACET project was a collaborative project investigating the potential of semantic expansion in retrieval. It aimed to take advantage of facet structure in both the interface and retrieval mechanism. More information and a web demonstrator can be found on the project webpages. This research strand is ongoing - see publications. We would be interested to hear from anyone who wishes to apply FACET query expansion techniques to other applications.
KOS - Ontology linking
KOS can serve as components of higher level ontologies. This is being pursued in two projects: 1) A collaboration with English Heritage exploiting the CIDOC Conceptual Reference Model. 2) A DELOS (Network of Excellence on Digital Libraries) JPA collaborative project, part of DELOS CLuster 5 on Knowledge Extraction and Semantic Interoperability. More details shortly.
KOS - Natural Language research
Linguistic techniques can assist problems such as the disambiguation of homographs when mapping to controlled terminology. Iolo Jones PhD research is investigating the potential of linguistic techniques for KOS-based indexing and searching.
Evaluation and user studies of KOS-based searching
Dorothee Blocks' PhD research investigated information searching behaviour with KOS-assisted systems and involved empirical study of FACET users and evaluation of different prototypes. A model of the KOS search process was devloped, with particlar reference to key areas of support in thesaurus use.
Terminology Services and tools
Embodied as web-based services, KOS can facilitate resource discovery and retrieval. We have developed an experimental web service, 'rich client' browser application, which can be downloaded. More information and a downloadable version here.
XML/RDF KOS representations
XML/RDF formats for the representation of KOS data, and demonstrations of the use of XPATH querying techniques.
We have developed some lightweight and flexible XML formats for the purposes of representing and storing thesaurus data for various research projects. As part of this work, we have experimented with XPATH querying techniques. As part of this work, w ehave done some iniitial experiments with the the Simple Knowledge Organisation System (SKOS) RDF format. Various schema files, demonstration data files and usage examples may be viewed or downloaded here.
NKOS Workshops
We have been involved in the organisation of Workshops on Networked KOS at ECDL (European Conference on Digital Libraries) in 2000, 2003, 2004 and 2005. A special issue of the Journal of Digital Information Systems (JoDI), co-edited by Tudhope & Koch, resulted from the 2004 workshop.
An NKOS Workshop took place on September 22, as part of ECDL 2005 in Vienna. Full details with presentations are on the workshop website. Note there is currently a cfp for an (early 2006) NKOS special issue of the journal New Review of Hypermedia and Multimedia on an NKOS theme.
Previous Projects
Geographical KOS and gazetteers
Many digital collections incorporate some form of spatial indexing and for many
users access via placename is important. Not all such collections require a full
GIS but may rely on a more limited spatial descriptor (footprint). Such
representations occur in online gazetteers, geographical thesauri or
ontology-based geographic applications where conventional GIS datasets may be
unavailable or pose undesirable bandwidth limitations. The Ontologically Augmented Spatial Information System (OASIS) project explored the use of spatial relationships to assist search with a placename component.
Semantic Hypermedia Architecture
The project builds on work that started in 1991 when the University was commissioned to develop a hypermedia museum exhibit on local history from the photographic archives of the Pontypridd Historical and Cultural Centre. This inspired the Semantic Hypermedia Architecture (SHA) project to investigate a query-based approach to hypermedia access and retrieval, rather than relying on fixed, embedded links. In that work, primary access routes were time, space and, as subject index, the Social History and Industrial Classification. Various research prototypes investigated a hypermedia architecture with a semantic index space separate from the document space.
| 
