In 2015, one year before the start of the CAS initiative, Richard Marciano and Michael Kurtz launched two pre-CAS pedagogical initiatives at the U. Maryland, the Digital Curation Innovation Center (DCIC) (https://ai-collaboratory.net/wp-content/uploads/2023/12/Brochure_2018-2019.pdf) and the Digital Curation for Information Professionals (DCIP) (https://info.umd.edu/academics/certificates-non-degree-study/digital-curation-for-information-professionals-certificate/).

The DCIC aimed to create a hands-on lab culture for LIS students, promoting interdisciplinary projects that explored the integration of archival research data and technology to generate new forms of analysis for historical research engagement. Projects were structured around social justice and infrastructure themes (https://ai-collaboratory.net/wp-content/uploads/2023/12/Project-Booklet_2018-2019.pdf). Student-led ‘datathons’ included an 8-week UMD Digital Curation sprint on Japanese-American WWII Camps (https://ai-collaboratory.net/projects/ct-ja_ww2_camps/digital-curation-students-and-filmmaker-event/) and a 2-day Maryland State Archives (MSA) datathon on Legacy of Slavery (https://ai-collaboratory.net/projects/ct-los/student-led-datathon-at-the-maryland-state-archives/). The DCIP was the U. Maryland iSchool’s first professional certificate program designed for individuals currently in the library, archives, or digital curation field, or who were planning to enter it but wish to enhance their digital skills. The program (2015-2024), through three online courses over 9 months, took students from introductory through advanced digital curation lessons while focusing on practical applications and a final hands-on capstone project. Examples of projects include 2021 (https://ai-collaboratory.net/2021/08/23/dcip-cohort-2021-presents-capstone-projects/), 2022 (https://ai-collaboratory.net/2022/09/07/2022-dcip-cohort-presents-their-capstone-projects/), and 2023 (https://ai-collaboratory.net/2023/08/14/2023-dcip-cohort-presents-their-capstone-projects/), and 2024 (https://ai-collaboratory.net/2024/09/03/2024-dcip-cohort-presents-their-capstone-projects/). In 2020, with the COVID pandemic, Marciano and Lemieux joined forces with colleagues and pivoted to an online virtual research network, the Advanced Information Collaboratory (AIC) (https://ai-collaboratory.net/). The focus is on: (1) exploring the opportunities and challenges of “disruptive technologies” for archives and records management (including digital curation, ML, AI), (2) leveraging the latest technologies to unlock the hidden information in massive stores of records, (3) pursuing multidisciplinary collaborations to share relevant knowledge across domains, (4) training current and future generations of information professionals to think computationally and rapidly adapt new technologies to meet their increasingly large and complex workloads, and (5) promoting ethical information access and use.

Computational activities are now being incorporated into regularly scheduled LIS graduate courses. Our goal is to modernize archival and library education and contribute to the development of faculty and library “digital leaders.” Examples include:

INST747: research in advanced digital curation (see Jun. 18, 2022, student project https://ai-collaboratory.net/2022/06/18/june-18-2022-measuring-the-impact-of-urban-renewal/)

Computational exposure was also incorporated into undergraduate Information Science courses. Examples include:

INST341: introduction to digital curation (see Dec. 9, 2021 (https://ai-collaboratory.net/2021/12/14/dec-9-2021-digital-curation-showcase/)).

Incorporating computational research and training has led to media outcomes for students and portfolio projects they can incorporate into their CV, such as the followings:

Terp magazine Winter 2023 – “truth in exile” (https://terp.umd.edu/truth-in-exile), with a project featured on ‘NHK World Japan television’ (https://ischool.umd.edu/news/aic-umd-student-project-featured-on-japanese-television/).

Increasingly, we are exposing students (LIS grad and undergrad) to data science and archives through the creation of Jupyter Notebooks. See student feedback (Piety, 2023), with presentations and notebooks (https://ai-collaboratory.net/2023/05/11/may-11-2023-computational-storytelling-datathon/). The ultimate goal is to contribute to the development of faculty and library digital leaders. In addition to preparing our students to meet these challenges through research-oriented seminars, we are building an online repository called CASES (Computational Archival Research Educational System) involving CAS practices that address practical digital records management and archival problems (https://cases.umd.edu/). We are exploring how computational thinking practices might be introduced to graduate students in the core Archival Studies curriculum. The goal is to enable a collaborative network of educators and practitioners who can learn from one another through the sharing and dissemination of computational case studies and lesson plans.

Teaching in CAS began at the University of British Columbia (UBC) in a course developed by Victoria Lemieux in 2011 on Information Visualization and Visual Analytics. The course, which began as a highly experimental topics course in its early days and continues to this day, now has a permanent place in UBC’s Masters of Archival Studies (MAS) and Master of Library and Information Science (MLIS) programs. Since the takeaways from this early CAS work for archival education have already been discussed (Marciano et al., 2018a), here it is sufficient to remark that the content of this course was one of the first archival courses to expose students to the application of machine learning, artificial intelligence, and data visualization to archival work and research, Through the subsequent writing of Devon Mordell, who was a student in one of the offerings of the course, it also was able to seed the formulation of a new ‘archives-as-data’ paradigm (Mordell, 2019).

Experimentation with CAS curriculum and pedagogy continued at UBC with the introduction of a summer institute on blockchain technology in 2017 and the formal introduction of a Blockchain Graduate Pathway in 2019. The program has been recognized by Coindesk as number one in Canada for blockchain education (Coindesk, 2023). In its current formulation, the blockchain graduate pathway aims to provide students with the knowledge and skills to contribute to the advancement of research and development in blockchain technology while at the same time engaging in some of the world’s most complex sociotechnical issues. The multidisciplinary blockchain graduate pathway comprises a 12-credit non-degree training program that augments existing master’s and PhD programs at UBC. Students in the pathway take courses on ‘issues and perspectives on blockchain technology’; technical ‘foundations in blockchain and distributed ledger technology’; blockchain and distributed ledger technology research; and ‘blockchain for information professionals’, which delves into various records and information-related aspects of blockchain technology, such as their nature as records and record keeping systems, archival theoretic critiques of blockchain solution design, and compliance with regulations, including privacy, data protection and copyright.

In addition to completing coursework, students complete a research-based industry internship and prepare a major research paper or thesis based on the research conducted for the internship. The following papers represent examples of outcomes of this work: (Lemieux, Hofman, Batista, Joo 2019; Kang 2021; Lemieux, Voskobojnikov & Kang 2021; Suleman 2022; Suleman & Lemieux, 2023).

MAS and MLIS students who have completed this program have gained knowledge and experience equipping them to fill a range of archival, records management, and data and information management roles, as attested to by this former student:

I first heard of the blockchain @ UBC program in the iSchool newsletter. Having had no prior understanding of blockchain beyond knowing it had something to do with Bitcoin, I was interested in knowing more. The program provided me with an interdisciplinary knowledge of blockchain technology and the countless industries and use cases it can be applied to. Through the course work, the Summer Institute, and an internship with the Land Title and Survey Authority of British Columbia, I was able to apply theoretical concepts in the real world. While I currently do not work within the blockchain space, the transferable skills I gained through the program allowed me to become an information specialist at a technology company a few months prior to graduation. (See: UBC iSchool, blockchain graduate pathway: https://ischool.ubc.ca/graduate/areas-of-focus/blockchainubc/).

The final example in this case study of integrating CAS into the curriculum for students of archives is a recently redesigned offering of a first-year core graduate course on archives on technology (ARST 500). The course aims to provide knowledge of the role of technology in archival work, at the theoretical and pragmatic level. Specific learning outcomes of this course are as follows:

Discuss the information and communication technologies in use in archives work and the affordances, both positive and negative, of these technologies

Themes of archives as data; comparative notions of integrity and authenticity; knowledge representation; power and positionality; and privacy are also explored in this course.

One of the primary methods by which the learning outcomes are achieved is by assigning students a series of labs designed to help them explore the application of technology in support of archival work, and the implications of doing so in concrete and practical terms. In the 2023 offering of the course, groups of students were asked to conduct Natural Language Processing and Sentiment Analysis on supplied sample text extracted from a 1920 newspaper. To complete the lab, they used Google Colabs, a free Jupyter Notebook environment that requires no setup and runs entirely in the cloud, which eliminated the need for students to download software or set up a virtual environment to run the lab. Google Colab was used to run the Transformers Library by Hugging Face using the DistilBERT model. Transformers offers an open-source library that provides state- of-the-art machine learning models, primarily for tasks in natural language processing (NLP). As students were not expected to have prior programming experience, and the pedagogical purpose of the lab was not to teach them programming but rather to expose them to processes of datafication of archives, how natural language processing works, and the implications of using such tools in relation to data/records integrity and bias, they were given a ‘recipe’ containing the script in Python to run the sentiment classifier rather than being asked to write the code themselves.

To stimulate students’ critical thinking on the application of sentiment analysis to archival material, the lab was paired with readings, including (Mordel, 2019) and (Jo & Gebru, 2020), which students were asked to connect back with their experience of conducting the lab.

Several groups went beyond the specific assigned task to conduct further experiments using different sentiment classifiers and text. For example, one student group ran an experiment using the SpaCy sentiment classifier on Chinese text, generally observing that the results were consistent with what the students would have expected the classification of the words to be. Another group ran the DistilBERT model over specific words relating to sexuality (see Figure 4), concluding that the results showed a ‘homophobic’ bias, and that the classifier was unable ‘to understand the multiplicity of meanings [of] a word…which is quite problematic’. This example of experiencing various flavours of sentiment analysis serves the purpose of developing critical thinking skills that show the strengths and weaknesses of computational techniques.

Successful examples of emerging educator networks can be found in Buchanan et al. 2023), where LIS students developed archival case studies with Jupyter Notebooks that document the International Research Portal for Records Related to Nazi-Era Cultural Property (IRP2), and a selection of unclassified catalogue entries about digitized nuclear science reports. Archival data visualizations are shown to promote outreach and access. Other examples include the LEADING network (LIS education and data science integrated network group) focused on preparing a cohort of LIS doctoral students and early to mid-career librarians for data science endeavours (Greenberg et al., 2023).

Beyond these emerging educator networks, to support the education of archival practitioners who are competent in CAS, it is necessary for there to be willing learners and archival educators to teach. This has been a challenge, as noted in a recent presentation made by Anne Gilliland on the theme of ‘preparing archivists in computational thinking and innovative technologies’ at the international academic conference on digital intelligence: empowering the modernization of archival work, hosted by the school of history and culture, Shandong University, China in October of 2023,

It has been quite difficult to even develop the teachers of archivists, the professors who teach them, to develop those skills or even the awareness of where computational processes would be appropriate to use in archival work, archival work, as opposed to digital humanities, work or STEM. and I think it has been equally difficult to persuade students that this is not just an important part of archival work but an essential part of archival work.

Our observation is that archival educators, by and large, lack the technical knowledge and skills to effectively teach CAS. Moreover, there is seldom institutional support to help them acquire the necessary knowledge and skills, as discussed in a recent paper by Lise Jaillant (2024). It takes incredible dedication and effort for someone ‘classically trained’ in archival studies to acquire the technical capability to teach CAS. Equally, it takes training for anyone ‘classically trained’ in technical disciplines, such as computer science or engineering, to acquire archival knowledge and skills to teach CAS. The CAS educational challenge lies not only in the absence of technical knowledge and skills, however, but also in a lack of technical curricular content and pedagogical knowledge of how to teach such material to learners with very different levels of technical knowledge, skill and aptitudes.

Perhaps a greater challenge is motivation, though: while the transformation in archival work wrought by computation should be motivation enough for archival educators to ‘grasp the nettle’, as Gilliland observes, this does not always happen. Similarly, more technical disciplines may lack interest in working with cultural materials and fail to appreciate the many other applications of archival theory (e.g., identifying deep fakes, to name but one ‘au courant’ area).

To address the shortage of CAS educators and the many challenges of providing CAS education, and to stimulate a meaningful dialogue among educators of diverse disciplinary backgrounds with a view to developing CAS as a novel trans-discipline, we are now engaging in building a TALENT Network for the Training of Archival & Library Educators with iNnovative Technologies). See: https://ai-collaboratory.net/projects/talent-network/. TALENT brings together experts from across the United States (including archivists, librarians, Library and Information Science educators, historians, learning scientists, cognitive scientists, computer scientists, and software engineers) to create a durable, diverse, and multidisciplinary national community focused on developing digital expertise and leadership skills among archival and library educators. The network focuses on four major objectives: (1) doubling the educator network to include diverse educators in multidisciplinary iSchools that have a focus on adjacent disciplines (e.g., computing, engineering, education, and data science) where there are substantial needs for archivists and librarians with digital and computational skills, (2) engaging HBCU students (Historically Black Colleges and Universities), (3) conducting curriculum development through experts from a Learning Sciences Network, and (4) addressing the social and ethical concerns that arise from computational and algorithmic thinking. See Figure 5.

All of the efforts to build and sustain CAS educator networks are for naught, of course, unless, as already mentioned, there are willing students. We note that in recent archival discourse, technology has often been associated with colonialism and other negatively perceived ‘isms’, as this statement exemplifies:

in 2011, the Pluralizing the Archival Curriculum Group (PACG) of the Archival Education and Research Institute noted that “Archival studies education programs are conceptualized in strikingly similar ways worldwide, largely because of the overarching bureaucratically- and legally-centred paradigms developed and exported from Europe through colonialism, evangelism, mercantilism, and technological developments, and later codified through national and international standards and terminologies (PACG, 2011).

Jaillant (2024) makes a similar observation about negative perceptions of technology. While efforts in pluralizing the archival curriculum are indisputably enriching, it is worth considering the potentially ‘chilling’ effect such perspectives may have on students’ willingness to learn about CAS and develop the technical skills that, as Gilliland (2023) and Jaillant (2024) observe, are not merely important but, nowadays, an essential part of archival work. This begs the question of how best to bring pluralizing perspectives into dialogue with CAS perspectives. Certainly, educational offerings focusing on the ethical aspects of applying technology to archival work, as discussed above, and those that exemplify how CAS techniques can be applied to advance social justice themes, help to achieve this end, but more effort is needed. The desire to advance this effort further is what, in part, motivated us to write this chapter, and also partly shaped our thinking on future directions for CAS, as we discuss in the following section.