ABSTRACT. This paper illustrates how to design and implement an engaged computational archival framework that leverages big archival records in order to respond to social justice and reparations policy imperatives. The work touches on two of the conference themes: (1) how to handle histories of people whose lives were deeply impacted by public authorities, and (2) Archives as Big Data as a potential restorative strategy.

Over the last few years Computational Archival Science (CAS) has emerged as a new discipline that explores the use and consequences of emerging methods and technologies around big data with archival practice and new forms of analysis and historical, social, scientific, and cultural research engagement with archives.

Our paper presents a very timely case study focusing on the legacy of urban renewal in Asheville, North Carolina between 1965 and 1980, when housing policies were enacted that ultimately displaced and erased African American businesses and communities with traumatic and lasting effects. "Urban Renewal was a program created by the U.S. Federal Housing Act of 1949, with the intention of redeveloping areas of cities that were deemed blighted." 

The study discusses making community members the focus of archives, and designing new interfaces to tell human stories. We explore CAS in the context of reparation, truth and reconciliation based on an earlier project developed by the U. Maryland team. On March 15, 2022 a Reparations Commission was finally formed, with ten seats for appointments representing the areas of criminal justice, economic development, education, health care, and housing, and fifteen seats for residents of historically impacted African American neighborhoods.

The authors of this paper believe this work serves as a model for other historical types of reparation that can benefit from CAS approaches.

***

PRESENTERS: Richard Marciano and Arthur Ray McCoy. 

ABSTRACT. Artificial Intelligence (AI) has recently gained relevance in the context of digital archiving practices on a global scale. However, multiple factors weigh against a smooth implementation of these technologies in memory institutions, such as metadata quality, affordability of professionals and adequate infrastructures. The case studies in the literature mostly represent projects of a certain size, funded by national or international schemes and concerning selected — sometimes unpublished — datasets that might not become publicly available once engineered with AI. The latter is a manifestation of the complex equilibrium between (meta)data FAIRness and openness due to copyrights and various ethical concerns. The available digitised collections present content discovery challenges, which may vary depending on users, an aspect that archival institutions might want to consider when making their collections accessible. Metadata quality is the main limiting factor for effective content retrieval and, despite the existence of metrics allowing quality assessment, metadata requirements for specific use and re-use cases (such as machine readability for AI tasks) are underexplored. The Europeana aggregator brings together more than 50 million digital Heritage objects from 3,500+ European Galleries; numbers that, combined with the presence of a certain extent of standardisation, make it a suitable candidate to investigate accessibility and content retrieval in the context of AI. Does Europeana support the discovery and retrieval of datasets suitable for AI tasks? Should archivists reconsider providing access to those who wish to re-work content with AI? Is AI implementation calling for a change in the current archival access practices? This work proposes a critical practice-based reflection on the implication of access and content retrieval through Europeana, specifically for Computer Vision (CV) tasks within the context of Irish historical photographs.

ABSTRACT. The Danish National Archives serve as the main repository of research data and other digital collections created by public institutions on a national, regional and local level. It is a core part of the archive’s institutional mission to secure the documentation of the Danish society’s past and present, and to facilitate that the data is used and reused to create value for society. It is an area of strategic focus that the collections are easily accessed by everyone, including scholars from various fields.[1] Considering the size of the archive’s digital collections, the scope of data re-use is currently limited. The growing awareness of this issue has prompted the development of systems for better access.[2] However, we would argue that a more thorough rethinking of the ways in which our data could be repurposed and suited to researchers’ needs is necessary. In this presentation, we aim to discuss how to move forward in thinking about data, ethics and the end-user, which can help us ensure a fruitful re-use of our digital collections. Moving beyond the classic approach of thinking only users in relation with our collections, we turn our attention to researchers and their environments, and the processes by which datasets become subjects of scholarly attention. This is not just a question about digitization, but also research practice. We propose three lenses to consider: 1. The researcher, who is defined by a several dimensions (field of interest, career stage, technical competence, attitude).[5,6] 2. The constraints of Western Academia (time pressure, institutional requirements of novelty and excellence).[4] 3. The ecosystem of digital resources that are already available, both nationally and globally,[8] and the user profiles according to the type of materials in our collections[9,10]. We will present the results of our ongoing discussions and future plans.

ABSTRACT. In 2005, a Swedish television documentary that revealed gross misconducts and abuse in Swedish childrens’ homes led to a number of inquiries that culminated with the passing of the Redress Act seven years later. This act entitled everyone that had suffered from abuse in out of home care to 250,000 Swedish crowns (close to 25,000 Euros). However, only 42 percent of the Swedish claimants were granted the compensation, a very low number in international comparison. This paper takes off in this – in many accounts failed - redress process. It investigates the role of records and archives in the redress scheme, and discusses how the concepts of total archives and politics of regret are present in and relevant for it. It also takes into account what impact narratives and narrativization had on the redress process. These questions are addressed by an inquiry into a few schemes of redress, mainly focusing on Sweden but also drawing comparison with Norway. Previous research has shown that a fragmentation of narratives contributed to the Swedish assessments being startlingly strict in an international comparison. In the present paper, this fragmentation of narrative is tied to the Swedish archival system and practices of records management. The findings highlight the impact by different archival strategies and practices on the lives of individuals and problematizes its connection to society.

ABSTRACT. This paper will focus on the value of metadata in relation to data reuse. Therefore, the problem in the analysis will be the challenges that one may face in connection with larger data projects with varying quality of data. The starting point for this paper will be the National Archives' recent work to build up a database for those applicants who want to apply for early retirement, the so-called Arne pension.

The crucial point was therefore whether the National Archives could provide documentation of the various forms of wage income up through the 1970s. In this process of clarification, the National Archives launched a major analysis of metadata, which formed the basis for the establishment of the largest collection of historical data to date. One of the crucial realizations in this process was that the value and quality of metadata is crucial for one to even consider the possibility of reusing data. This paper will therefore address and analyze some of the specific issues that the National Archives faced in 2021. The assumption is that these issues will contain a bid for all archives that intend to reuse data.

The research question in this paper will therefore be; How will the National Archives solve the public / society's need for historical data / documentation in the future, when metadata is either absent or deficient? What can the archives do in the future so that their historical data can be used for analysis in relation to relevant current issues such as pensions, health / epidemics, climate / energy policy, data for analysis and policy development, and major research projects in personal and register data using new analysis technologies?