Current good practice in search and discovery: your help invited

With a view to informing the search and discovery strategy for the Georgian Papers Programme, Chris Olver, Metadata Coordinator for the GPP at King’s College London,  has surveyed over 40 historical sites variously developed in the UK, USA, Australia, New Zealand Canada, Germany and. The Netherlands The survey included examples of manuscript transcription projects, historical databases, meta-aggregators, on line finding aids and electronic printed editions. A majority of the sites surveyed included a significant quantity of eighteenth century material but good practice in subject indexing, use of authority files and linked data was also of interest where it was likely that approaches were applicable to the Programme. Sites had variously been compiled by teams of academics, librarians and archivists and some predated the internet.

All suggestions about well-regarded and used sites have been followed up and further suggestions and thoughts will be very welcome. Do please tell us about which sites you find useful and why or indeed what you would like to do on a site but can’t.

Below is a list of the sites surveyed to date:

This slideshow requires JavaScript.

Our conclusions to date from the survey about what makes a good site are:

– Websites need to be simple to use and easy to navigate and allow users to filter results in a number of different ways.

– The most effective browsing options navigate through authority files.
– The use of international standards based authority files facilitates cross searching other databases.
– Subject headings (hierarchically arranged) provide more focused search terms than keywords.

– The provision of open access to datasets allows researchers the ability to download records and preform textual and visual analyses that may not be possible within the site.

To these we would add some of the key conclusions of work undertaken in developing the Wellcome Medical Heritage Library:

– The need to provide an explanation of the corpus- what is and what is not on or covered by the site
– Sites should be capable of revealing networks of people, places, practices, suppliers, thoughts etc.
– Browsing/discovery through serendipity remains essential-the lightbulb idea that springs from chance discovery
– The integrity of the digital object must be preserved. Wellcome describe this as the ‘bookishness of the book’. For archivists this is about preserving the provenance of the object and associated contextual information.

These ideas were shared at the recent Omohundro Institute for Early American History and Culture annual conference in Worcester, MA and valuable feedback obtained. Colleagues in attendance suggested the following additional sites were important for their work:

  • Brieven als buit [Letters are Loot]: research database constructed by Leiden University allowing access to 40,000 Dutch letters from 17-early 19th century gathered from British archives. The website is extremely well indexed and allows visitors to search using Corpus Query language.
  • Documenting the American South: DocSouth provides access to a range of digital records of the American South run by the University of North Carolina. Site is very well indexed with multiple search pathways including browsing by Libary of Congress Subject Headings. 
  • Empire Online: a digital repository created by Adam Matthews, publishers. The site was recommended for having excellent contextual essays relating to the site content.
  • Franke-halle.de: The digital library of the Franke Foundations was recommended for having great search features, including faced navigation and indexing. The library also has incorporated an adjustable document viewer and includes full text transcriptions.
  • Gallica (National Library of France): The Digital Library of the National Library of France, originally created in 1997 but recently overhauled, has a very impressive interface with multiple search pathways and a very sophisticated image viewer. 
  • Marine Lives: provides substantial amount of transcribed mid-17th century content from the English Admiralty Court. 
  • Massachusetts Historical Society: The oldest historical society in USA has a vast collection of digital resources, teaching aids, blogs and current events. 
  • MEAD (The Magazine for Early American Datasets): This online repository for historians’ datasets is a fantastic concept. The datasets are free to download with the metadata capturing donor information, context of work and place of origin. 
  • Nines (Networked Infrastructure for Nineteenth-Century Electronic Scholarship): is a meta-aggregator that collects information from 139 nineteenth century sources.  
  • Trans-Atlantic Slave Trade Database: remarkable series of databases that was originally published on CD-ROM in 1999. The project has been running for decades and provides information on voyages undertaken, estimates for numbers of slaves and an African names database. 

We will be following up these suggestions and look forward to hearing your thoughts on search and discovery.

Patricia Methven, GPP Programme Manager, King’s College London

Leave a Reply