Eliminating query failures in a work-centric library meta-search environment

Abstract

Purpose - This paper aims to describe how approaches such as semantic based query rewritings and FRBR work entities composition could solve existing problems and improve the overall behavior of a mediated based meta-search environment. Moreover, it aims to present meta-Composer, which is a meta-search engine without query failures that composes work level entities for display, as well as an autonomous open service for discovering substitutions of unsupported access points, in the context of the Z39.50 environment. Design-methodology-approach - Semantic Access Point Network (SAPN) engine, which is a system for substitution of unsupported access points by exploiting semantic access point information from an RDFS description, was extended as a query rewriting system taking into account user preferences for recall and precision. Additionally, the Library of Congress FRBR display tool was extended to create, in the first place, MARC21, UNIMARC and MODS wrappers for the production of work-expression-manifestation identifiers, and, secondly, the composition of work clusters. Finally, pazpar2, the open source meta-search engine was used as both the middleware and the client platform for the integration of all the above components and the development of meta-Composer. Findings - Consistent extension of the meta-search search functionality is possible from the lowest common denominator of the supported access points through semantic based query rewritings. Furthermore, enabling query specificity, which restricts the resulting records, could really improve the overall performance of the system and allow for more space for other extensions. Work centric result presentation gives a more compact and coherent view to the user compared to the typical record-based de-duplicated display. Research limitations-implications - The technical limitation for retrieving all the results from all sources, especially for large result sets, indicates the need for further research for the development of heuristics in order to find the best sources for enriching the results by unique work clusters. The absence of tools, such as authority files, limits the accuracy of the FRBRization process and the discovery of correlations between related works. Practical implications - Consistent meta-search services require widespread use of international metadata standards and communication protocols. Originality-value - The SAPN technology enforces implicitly the semantic interoperability in the context of the Z39.50 protocol. meta-Composer is an open service which offers meta-search within valuable sources world wide and it also provides the potential to evaluate new services. The SAPN open service gives the ability to any other mediator or client, to extend and improve consistently its search functionality, in the same context and by using web services technology.