Category: Resource

Kelda Habing and Lian Ruan (University of Illinois Fire Service Institute) have published a qualitative study on digital humanities practices from seven US academic libraries to provide insights into how varied academic libraries operate their DH programs, in Digital Transformation and Society. Using semi-structured interviews, they highlight practices around space, technology, staffing, instruction, and collaboration. Their article contributes to the growing corpora of literature and studies addressing trends and approaches in academic libraries for offering DH services and support. From the abstract:

Purpose
Digital Humanities is a robust area of research and practice at universities and their libraries across the world. This case study investigates the unique DH practices of seven US academic libraries to provide insights into how varied academic libraries operate their DH programs.

Design/methodology/approach
Semi-structured interviews with nine library staff in DH or DH-adjacent positions at seven US academic libraries were used to investigate library DH practices.

Findings
This case study highlighted key areas of academic library DH practices including Space, Technology, Staff, Instruction and Collaboration. Practices in these areas were compared against each other and literature to comment on the current state of DH library practices and offer some recommendations for select areas.

Research limitations/implications
This case study interviewed staff in a limited number of US libraries and is not generalizable to or a reflection of the many academic libraries in the US or across the world.

Originality/value
The juxtaposition of multiple libraries’ DH activities provides a unique perspective on academic library DH practice, as many studies investigate only a single library as their subject.

 

The Association of Research Libraries (ARL) has published “Research Libraries Guiding Principals for Artificial Intelligence,” a brief values-statement on the use of AI in their policy advocacy and engagement. From the background statement:

Articulating a set of research library guiding principles for AI is useful to influence policy and advocate for the responsible development and deployment of AI technologies, promote ethical and transparent practices, and build trust among stakeholders, within research libraries as well as across the research environment. These principles will serve as a foundational framework for the ethical and transparent use of AI and reflect the values we hold in research libraries. ARL will rely on these principles in our policy advocacy and engagement.

The document, available via html or pdf download, consists of 7 guiding principles, from “Libraries democratize access to AI tools and technology to foster digital literacy among all people,” to “Libraries negotiate to preserve the scholarly use of digital information.” The document contains linked references to additional resources on AI in research libraries.

This week’s editors-at-large nominated “Exploring LLM Weirdness” by Cord, a quiz that requires players to convince Chat-GPT4 to select the right answer in a multiple choice quiz. It serves as an interactive lesson and teaching tool on the limitations of AI in certain scenarios.

The latest lesson from Programming Historian is “Working with Named Places: How and Why to Build a Gazetteer” by Susan Grunewald and Ruth Mostern. The lesson takes the learner through the process of creating a gazetteer from historical texts and then shows how one might leverage the gazetteer’s data using linked open data and GIS. A notable and welcome feature is the focus on concepts of place to give the learner a solid foundation and a deeper understanding of gazetteers as “a kind of Knowledge Organization System (KOS), a tool ‘that brings together related concepts and their names in a meaningful way, such that users of the KOS can easily comprehend the relationships represented’.”

The collaborative, NEH-funded AI for Humanists project (formerly the BERT for Humanists project) creates learning resources aimed at empowering humanities scholars to use machine learning and artificial intelligence tools, specifically large language models (LLMs) in creative new ways.

Adding to its repository of tutorials, primarily for python coding, two new tutorials have recently been published on the project website:

• Measuring Document Similarity with LLMs: “This code notebook demonstrates how you can use LLMs to explore which texts, or documents, are similar to each other in a given dataset. We explore narrative vs. non-narrative texts, historical poetry, and ChatGPT-generated poetry.”
• Zero-shot Prompting with LLMs: “In this tutorial, we specifically explore how you can prompt a model to predict the genre of a book based on its Goodreads review and to predict whether a given passage is narrative or non-narrative text. But you should be able to use and modify this workflow for your own text classification needs.”

Quinn Dombrowski has updated A guide to writing DH conference submissions with How to Write an ADHO DH Conference Proposal in 2023

I’ve been the ACH representative to ADHO since summer 2021, and one of the major initiatives that ADHO has undertaken has been to rework the review criteria for the conference. … The goal of the new review criteria was to elicit constructive feedback and not just negative comments, and functionally make reviewers justify their numeric choices through their text comments.

Dombrowski breaks down the structure of a submission based on the revised criteria into four categories: disciplinary context, DH context, detailed description, and stakes. Very useful advice for anyone planning to submit a proposal for DH2024 in Arlington, VA (deadline for proposals is December 5th!).

Ethics in Linked Data, edited by Alexandra Provo (New York University), Kathleen Burlingame (University of Pennsylvania), and B.M. Watson (University of British Columbia), “brings together contributions that explore ethics in linked data initiatives.” The book focuses on the idea that ethics should be at the forefront of consideration when the data is created, pre-existing damage should be acknowledged and mitigated, and more ethical outcomes should be a focus when working on linked data projects.  From the description:

Contributions investigate the intersection of linked data with such topics as gender, indigenous knowledge, inclusive data creation, authority control, identity management, systems design, codes of ethics, sustainability, critiques of fundamental linked data models, and more.

The Programming Historian recently published a new lesson, Clustering and Visualising Documents using Word Embeddings. Developed by Jonathan Reades and Jennie Williams, this lesson “uses word embeddings and clustering algorithms in Python to identify groups of similar documents in a corpus of approximately 9,000 academic abstracts. It will teach you the basics of dimensionality reduction for extracting structure from a large corpus and how to evaluate your results.”

Part of a special series in partnership with Jisc and The National Archives, the lesson includes background information, a case study, and instructions in dimensional reduction, hierarchical clustering, validation, and a bibliography that includes other relevant tutorials. It’s listed as high difficulty.

Given the ongoing deterioration of Twitter as a viable space for the sharing of digital humanities work, the Association for Computers and the Humanities has begun exploring alternatives. Bluesky social has the makings of a viable alternative, but it requires an invitation to join. This form was created to facilitate the sharing of these invitation codes. Once you’re on the platform you receive one of these codes every two weeks, so be sure to check back and pay it forward. Emails are only being used for the purposes of code sharing and recognition from ACH – they will not be made public.