This is a guest post by Annalina Caputo, Assistant Professor in the School of Computing, Dublin City University. She is working on the AURA Project with Lise Jaillant at Loughborough University as well as colleagues from National University of Ireland Galway, Waterford Institute of Technology and The National Archives.
AI is leading the fourth industrial revolution: its enhanced capabilities propelled by large access to big data and more powerful machines are changing the way we think about day-to-day tasks, machines, and society. Far from the fuss of big news titles, which promise new algorithms for driving cars, writing articles, and generating art masterpieces, the real “Intelligent Revolution” is already happening under our eyes, in a more pervasive and subtle way.
How will this Intelligent Revolution transform born-digital archives? What challenges and opportunities does it open for archives? Which ethical and legal aspects does it imply?
The AHRC-IRC research networking project AURA: Archives in the UK/Republic of Ireland & AI will try to answer these, and many other questions.
While the potential of AI has started to be unleashed in many aspects of our life, the application of AI to archives is still at an exploratory stage. In the past, much emphasis has been posed on digital preservation, but we now need to increasingly focus on access and usability. AI can not only facilitate access to documents, but it can also support the discovery and production of new knowledge.
Access to digital archives can be enhanced via:
- Representation: semantic representation of documents and artifacts will improve retrieval and exploration of archives, supporting the user interactions with the system in more meaningful and seamless ways. New algorithms able to extract and combine information from structured and unstructured documents will provide new levels of knowledge and discovery. Semantic web technologies and ontologies can help to represent levels of document/text confidentiality and support the automatic management of clearance to access them.
- Visualisation: new visualisation techniques and metaphors will support the users during the exploration of born-digital collections, and archivists/policymakers in their decision-making processes.
- Dialogue: New interaction paradigms based on natural language conversations rather than through fixed search boxes, will enable the access and use of digital archives far from the standard setting of a desktop, paving the road for more natural use of these collections in virtual and augmented reality.
The adoption of AI has set in motion a new realm of opportunities that do not come short of pitfalls. Bias, privacy and security issues, as well as lack of transparency and explanations, are some of the foreseeable risks involved with AI – in addition to all the unforeseeable problems that usually arise from the introduction of new technologies.
To increase trust in AI and collaborations between disciplines that are so different and distant, such as computer science and humanities, it is then important to set up a dialogue that overcomes technological and cultural gaps.
The AURA network will bridge the gap between Computer Science, the Humanities, and other stakeholders, to improve discoverability, accessibility, and use of born-digital archives.
The networking activities will advance knowledge and understanding of these three main fields.
- For Humanities scholars and archivists, these activities will shed light on new opportunities for ground-breaking solutions to the problem of born-digital archives.
- For computer scientists, they will provide rich and complex scenarios for applying theories to real-world problems.
- The collaboration and communication with the other non-academic stakeholders (archivists and policy makers) will provide insights and new perspectives on the problems of born-digital archives.
Over a period of 11 months, the AURA project will bring together a wide pool of people, spanning geographical and cross-discipline boundaries.
Co-led by Dr Lise Jaillant (Loughborough University), myself (Dublin City University), Prof. Mathieu d’Aquin (National University of Ireland Galway), Dr Larry Stapleton (Waterford Institute of Technology) and Dr Eirini Goudarouli (The National Archives), the project will foster collaborations between academics from the UK and Ireland, and non-academics project partners, such as The National Library of Ireland, The National Library of Scotland, the British Library, the Irish Traditional Music Archive, and the International Council on Archives.
The network themes will focus on three main topics that serve also as the centre of dialogue and debate during three workshops:
- Workshop 1 – 16th-17th November 2020 (Dublin, IE) will address the problems of “Openness versus Privacy” & “Issues of access born-digital archives”. In this two-day workshop, we will investigate problems, technical challenges, and possible solutions related to open data, privacy, and access of born-digital archives. Originally planned as a face-to-face event featuring a tour of the National Library of Ireland, this event has migrated online with the tour being now given virtually due to COVID-19 restrictions.
- Workshop 2 – 28th-29th January 2021 (London, UK) on “AI and Archives: Current Challenges and Prospects of Born-digital archives” will address the specific challenges and opportunities raised by the application of AI to archives.
- Workshop 3 – 16th March 2021 (Edinburgh, UK) will aim to outline and anticipate the impact and new landscapes opened by AI in "Artificial Intelligence and Archives: What comes next?"
The workshop will provide a virtual and physical space for experts and stakeholders, who rarely have opportunities, to meet and work together through a carefully designed structure of short presentations, roundtables, panels, and practical activities designed to lead to mutually beneficial, sustainable collaborations.
The conversation between parties will not be relegated to the space and time of the workshops, but it will continue, thanks to the network, in other forms of collaborations, through the editing of two special issues addressing computer scientists and humanities scholars, and publications of a policy paper and articles.
To increase participation and inclusion in the network, workshop recordings will be made available through the AURA network website (https://www.aura-network.net/).
To receive updates on our intitiative and latest news, subscribe to our mailing list (link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A0=AURA)