High-Level Thesaurus Phase IV: Transition to Service Testbed and Future Requirements Study
HILT
April 2007 – February 2009; additional funding to test embedding of HILT has extended the project to May 2009
HILT aims to ensure that FE and HE users of the JISC Information Environment can find appropriate learning, research and information resources by subject search and browse. In a distributed services environment where most service providers use different subject schemes to describe their resources such details of which are largely unknown to most users. The assumption is that services offering users facilities to cross-search and browse distributed services will use HILT services in a fashion transparent to their users. This is done via machine to machine (M2M) or web services protocols such as SOAP or SRW, through routines embedded in service user interfaces. On request, HILT will serve up data on individual subject schemes (broader terms, narrower terms, hierarchy information, preferred and non-preferred terms, and so on) and interoperability data (usually intellectual or automated mappings between schemes, but the architecture allows for the use of other methods). These data can be used by the information services to aid users in a variety of ways - for example, improving recall by enriching the set of terms known to a user by providing synonyms and related terms such as improving precision by providing the best terms for a subject search in a remote service that uses LCSH or UNESCO.
HILT Phase III built various pilot facilities allowing M2M interaction and also mapped out a design for an entry-level M2M service. HILT Phase IV is a transition to service phase that will allow this initial entry-level service to be built, tested for user requirements and retrieval effectiveness, refined in line with findings, and extended to permit the use of a range of distributed terminology services for interoperability. It will also involve a survey of the level of need and interest amongst JISC services in respect of an operational service and, if appropriate, a scoping of the cost and specification of a future operational phase of the service. A parallel programme of research into selected topics will help inform both the cost and requirements of an initial entry-level operational service and any future extension.
EDINA is providing programming support for development of the SRW client; advice on performance and interface issues; hosting the SRW server; and will be testing developments using a Go-Geo! client.
Centre for Digital Library Research, Strathclyde University