Serendipity in the digital humanities
Nostalgia for the alleged serendipity of the physical library remains a persistent (though disputed) trope in the digital era. There have even been attempts to mimic this kind of “accidental discovery” in the context of digital library services, as well as in search engines more broadly. Regardless of the merits of these initiatives, there seems to be a general acknowledgement that serendipity plays a significant role for humanities researchers – or at least that their research tends to incorporate the making of what appear to be accidental or fortuitous connections between apparently unrelated phenomena. In a digital environment dominated by the single search box, how can this kind of approach be encouraged and designed for?
This presentation will look at serendipity in the context of some of the thinking behind the Humanities Networked Infrastructure (HuNI) service (http://huni.net.au ). HuNI is designed to aggregate data from over thirty cultural datasets produced in Australia. They cover fields as varied as literature, history, indigenous studies, cinema and media studies, art and design, and linguistics. More than 730,000 named entities have been extracted from these datasets and made available for HuNI users to explore.
While HuNI provides discovery tools for casual users from the wider community, more sophisticated functionality is available to researchers who register for an account. They can authenticate themselves using social media logins and can share their discoveries and activities through their existing social media accounts. They are given their own personal workspace where they can create, save and publish selections of data, and can export data for reuse in external environments. They can assemble data into personal “collections” which can be made public for other users.
Researchers can also assert relationships between entity records in the form of ‘socially linked’ data, which are visualized as a dynamic network graph. These links become part of the public content of HuNI, and are available to all HuNI users for browsing. This linking capability contributes to the building of a ‘vernacular’ network of associations between HuNI records. It recognizes that there are diverse perspectives on knowledge and offers avenues for discovery which go beyond keyword and phrase searches.
HuNI is intended to support the kind of non-linear and associative research methods practiced in the humanities. It aims to encourage structured serendipity by enabling users to create their own links and associations in the data, within a loose framework of categorization. This approach is designed to minimize the imposition of a centralized or normative view of the data, while avoiding complete reliance on search algorithms.
Toby Burrows holds a European Union Marie Curie fellowship in the Department of Digital Humanities at King’s College London. He previously worked at the University of Western Australia, where his most recent role was as the manager of the eResearch Support Unit in the University’s Information Services Division. He has been involved in a series of Australian projects to develop digital infrastructure for humanities researchers, notably the national Humanities Networked Infrastructure (HuNI). He has a PhD in medieval history from the University of Western Australia and has previously held visiting fellowships at Churchill College Cambridge and the Free University in Amsterdam.