Endeca at the NCSU Libraries
Endeca-related Research Project Proposals
The implementation of Endeca's Information Access Platform to provide keyword searching for the NCSU Libraries catalog has generated a great deal
of interest in the future of library catalogs and in faceted navigation in particular. The Endeca Product Team at NCSU Libraries has developed
a list of research projects that relate to these areas. We welcome outside collaboration in exploring these topics (or others) of interest to the greater library community.
If you are interested in participating in a collaborative research project, please contact Emily Lynema or Andrew Pace.
Evaluation of Faceted Catalogs
Faceted navigation has grown in popularity over the past several years, with both libraries and vendors working to develop this type of search functionality.
However, open questions remain as to the effectiveness and usability of facets for different user populations and different user tasks.
- What is a good methodology for evaluating whether changes in the catalog truly improve its effectiveness / user experience?
- Is faceted navigation a better interface?
- When (and for who) is faceted navigation effective or useful? When is it not?
[top]
FRBR-ization in the Library Catalog
FRBR promises to reduce the complexity of search results by providing the ability to collocate catalog records that represent multiple versions of the same work. However, there are few
FRBR-ized search systems in production that provide a model for displaying results where aggregated work displays with multiple manifestations intermingle
with work displays represented by a single manifestation.
- How to design an effective user interface?
- How to analyze effectiveness of algorithms for aggregating manifestations?
[top]
Subject Access via Controlled Headings
Facets populated with subject headings help expose controlled subject access points to users
in a way that requires no prior knowledge of terminology. However, the use of controlled vocabulary to narrow a keyword search fails to retrieve
titles that do not contain the original keyword terms. In addition, the value of cross-references in leading users to appropriate subject terminology is lost in keyword searching.
- How to create an effective natural language entry vocabulary for LCSH?
- How to lead users from natural language searches to appropriate LCSH when searching?
Read a more detailed writeup of this research project.
[top]
Relevance Ranking for Bibliographic Records
Much research has been done in the field of information retrieval studying relevance ranking methodologies for full-text searching.
Little research has been done studying relevance ranking methodologies for metadata searching. The Melvyl Recommender Project at the California Digital Library
and the National Library of Australia (paper) are 2 institutions that have begun to explore this area.
- What is a good methodology for evaluating the effectiveness of changes in relevance ranking?
- What algorithms are most effective for ranking bibliographic records (metadata) as opposed to full text documents?
[top]
Enhanced Browsing Capabilities
The NCSU Libraries catalog provides a fairly simple implementation of facets where a user generally selects one value from each facet to refine a search.
It is possible to create a more powerful environment where users could select multiple values from each facet, combining these values to broaden (OR) or narrow (AND) a search.
Users can also be enabled to select facet values without submitting any search terms (browse).
- Is it possible to create an interface focused on exploration and discovery (via enhanced browse options), rather than known-item searching? Would such an interface be useful?
- Are there particular facets that would be more effective if multiple facet values were combined using OR?
[top]
|
