This year the annual iPRES digital preservation conference was understandably postponed and in its place the community hosted a 3-day Zoom conference called #WeMissiPRES. As two of the Bodleian Libraries’ Graduate Trainee Digital Archivists, Simon and I were in attendance and blogged about our experiences. This post contains some of my highlights.
The conference kicked off with a keynote by Geert Lovink. Geert is the founding director of the Institute of Network Cultures and the author of several books on critical Internet studies. His talk was wide-ranging and covered topics from the rise of so-called ‘Zoom fatigue’ (I guarantee you know this feeling by now) to how social media platforms affect all aspects of contemporary life, often in negative ways. Geert highlighted the importance of preserving social media in order to allow future generations to be able to understand the present historical moment. However, this is a complicated area of digital preservation because archiving social media presents a host of ethical and technical challenges. For instance, how do we accurately capture the experience of using social media when the content displayed to you is largely dictated by an algorithm that is not made public for us to replicate?
After the keynote I attended a series of talks about the ARCHIVER project. João Fernandes from CERN explained that the goal of this project is to improve archiving and digital preservation services for scientific and research data. Preservation solutions for this type of data need to be cost-effective, scalable, and capable of ingesting amounts of data within the petabyte range. There were several further talks from companies who are submitting to the design phase of this project, including Matthew Addis from Arkivum. Matthew’s talk focused on the ways that digital preservation can be conducted on the industrial scale required to meet the brief and explained that Arkivum is collaborating with Google to achieve this, because Google’s cloud infrastructure can be leveraged for petabyte-scale storage. He also noted that while the marriage of preserved content with robust metadata is important in any digital preservation context, it is essential for repositories dealing with very complex scientific data.
In the afternoon I attended a range of talks that addressed new standards and technologies in digital preservation. Linas Cepinskas (Data Archiving and Networked Services (DANS)) spoke about a self-assessment tool for the FAIR principles, which is designed to assess whether data is Findable, Accessible, Interoperable and Reusable. Later, Barbara Sierman (DigitalPreservation.nl) and Ingrid Dillo (DANS) spoke about TRUST, a new set of guiding principles that are designed to map well with FAIR and assess the reliability of data repositories. Antonio Guillermo Martinez (LIBNOVA) gave a talk about his research into Artificial Intelligence and machine learning applied to digital preservation. Through case studies, he identified that AI is especially good at tasks such as anomaly detection and automatic metadata generation. However, he found that regardless of how well the AI performs, it needs to generate better explanations for its decisions, because it’s hard for human beings to build trust in automated decisions that we find opaque.
Paul Stokes from Jisc3C gave a talk on calculating the carbon costs of digital curation and unfortunately concluded that not much research has been done in this area. The need to improve the environmental sustainability of all human activity could not be more pressing and digital preservation is no exception, as approximately 3% of the world’s electricity is used by data centres. Paul also offered the statistic that enough power is consumed by data centres worldwide to boil 10,400,000,000,000,000 kettles – which is the most important digital preservation metric I can think of.
This conference was challenging and eye-opening because it gave me an insight into (complicated!) areas of digital preservation that I was not familiar with, particularly surrounding the challenges of preserving large quantities of scientific and research data. I’m very grateful to the speakers for sharing their research and to the organisers, who did a fantastic job of bringing the community together to bridge the gap between 2019 and 2021!