Description
Keynote presented at the iPres2024 conference in Gent, Belgium on September 18th 2024. The overall topic of the presentation is personal web archiving.
Former Farmer Chases Traces
Herbert Van de Sompel is an Honorary Fellow at Data Archiving and Networked Services (DANS) in The Netherlands and Visiting Professor at the Internet Technology & Data Science Lab of Ghent University. Throughout his career, Herbert has focused on infrastructure (tools, standards, protocols) to advance open access to scholarly communication and to enable long term archiving in a globally networked environment. He has made a wide range of contributions, including the Open Archives Initiative specifications (OAI-PMH, OAI-ORE, ResourceSync), the OpenURL reference linking framework, Web Annotation, Memento, Robust Links, COAR Notify, and the Signposting the Scholarly Web specifications. He was awarded the 2017 Paul Evan Peters Award for his lasting achievements in the creation and innovative use of network-based information resources and services to advance scholarship and intellectual productivity.
One of Herbert's key projects is Memento, a protocol and associated infrastructure that provides uniform access to archived webpages by means of their original URL, irrespective of the web archive that hosts them. With his Research & Prototyping Team at the Los Alamos National Laboratory, and in collaboration with colleagues from Old Dominion University, Herbert developed the protocol that uses the URL of an existing or vanished web page and a datetime to retrieve a prior version of the webpage, using an approach that is similar to how language or region-locked versions of webpages are retrieved. Many web archives provide a glimpse into now-deleted webpages that they have salvaged. With Memento as an enabler, navigation across these distributed web archives is possible. Moreover, Robust Links, which leverages Memento, can prevent reference rot by ensuring that the referenced webpages are archived in a timely manner and that references themselves support visiting both the archived and current webpage versions. As such, web references stay true to the citing author's original intent while also allowing to assess how pages have evolved since they were cited.
“It makes archived web materials easier accessible, and I am not the only one to believe that access is a necessary condition for preservation, and so in that way, Memento kind of helps to preserve web materials.”
Interview with Herbert van De Sompel - DPC 2010
A major motivation for Herbert's work - rooted in his early career as head of library automation at Ghent University - has been a concern with the ever growing consolidation and monopolization of the global information infrastructure, which is apparent in the web, generally, and in research communication, specifically. And, while centralization is not harmful by definition, parties that have achieved a monopoly position can rarely avoid the temptation to abuse it. For many years now, with likeminded colleagues, Herbert has been advocating for more decentralized approaches in which interoperability provides the glue to realize a system in which communicating research instead of enriching shareholders is central. Putting scholarly communication back into the hands of the researchers is essential as is infrastructure that allows libraries and archives to preserve the scholarly record for the long term.
In addition to working at Ghent University and the Los Alamos National Laboratory, Herbert has held positions as Visiting Professor in Computer Science at Cornell University, Director of e-Strategy and Programmes at the British Library, and Chief Innovation Officer at DANS. His work has, among others, been mentioned in the New Yorker, Nature, The Digital Preservation Newsletter, and New Scientist.