Built on CollectionBuilder
This site is a modification of CollectionBuilder, an open source framework for building digital collection websites. CollectionBuilder is developed by faculty librarians at the University of Idaho Library and the Center for Digital Inquiry and Learning, with support from the Institute of Museum and Library Services and the National Endowment for the Humanities.
CollectionBuilder was designed for galleries, libraries, archives, and museums. The things it expects are photographs, postcards, oral histories, and maps, described in a spreadsheet and turned into a browsable site by Jekyll. It is metadata driven, which means the navigation, the visualizations, and the search are all generated from how the items are described rather than hand built.
This modification treats journal articles as a collection to help organize and present my scholarly work.
- The metadata schema was rebuilt for scholarship. Fields such as source, format, and rights statement gave way to output type, venue, venue type, publication status, theme, sub-theme, shared trajectory, DOI, arXiv identifier, and citation count.
- Browse facets became Theme, Sub-theme, Trajectory, Output Type, Venue Type, Topic, and Author.
- The search index covers venue, theme, trajectory, and abstracts alongside the standard fields.
- Abstracts and citation counts come from OpenAlex, matched by DOI.
Four changes needed new code, each addressing something a digital collection does not need but a research catalogue does.
- Google Scholar indexing. Stock CollectionBuilder emits Dublin Core, Open Graph, and schema.org markup, none of which Google Scholar reads. A new include emits Highwire Press
citation_*tags, varying by output type so that a preprint is not filed as a journal article. - Scholarly citation. The museum style attribution block was replaced with a formatted reference, a generated BibTeX record, and a copy to clipboard control.
- A topic network. The stock tag cloud became a force directed graph in which two topics are joined when they appear on the same publication. Node sizes still follow CollectionBuilder’s own tag cloud scale.
- A publication timeline. The stock TimelineJS page was replaced with a custom timeline that places each publication in its year and colours it by sub-theme.
Data
Everything behind this site can be downloaded from the Data page as CSV and JSON, including a Frictionless Data package descriptor. That is stock CollectionBuilder behaviour, and it is one of the better arguments for the framework in a research setting. The catalogue is queryable, not just readable.
Technical Credits - CollectionBuilder
This digital collection is built with CollectionBuilder, an open source framework for creating digital collection and exhibit websites that is developed by faculty librarians at the University of Idaho Library following the Lib-Static methodology.
Using the CollectionBuilder-CSV template and the static website generator Jekyll, this project creates an engaging interface to explore driven by metadata.
