Since 2015, the International Image Interoperability Framework (IIIF) has implanted itself as a standard for the storage, sharing and manipulation of digital documents in the GLAM sector. In this panel, we shall hear from IIIF specialists and researchers from the DH community about how IIIF is used for research.

As the digital humanities have matured, the field increasingly calls for support of existing work in danger of obsolescence. This panel offers multiple perspectives on sustainability of digital projects, as well as their underlying data and infrastructure. Panelist presentations include concrete examples and discussion of the funding landscape.

The paper explores the role of maps as epistemic tools in the Stadt.Geschichte.Basel project, emphasizing how spatial data dynamics and moments of disconcertment foster interdisciplinary collaboration. By embracing ambiguity and conflict in map-making, the authors create accessible, inclusive outputs that reimagine historical narratives and advance participatory scholarship.

This study employs digital humanities methods and hierarchical hexagonal spatial indexing (H3) to analyze patterns of police violence against African Americans across the United States. Using Local Moran's I statistics on over 13,000 incidents between 2015-2024, it identifies significant geographic clusters and transition zones, revealing how policing practices vary across jurisdictional boundaries.

This paper presents a case study of a data visualization involving the Trans-Atlantic Slave Trade Database. We explain how theories about historical trauma and the limits of recovery guided our work. We describe our design process, and propose a series of questions that can guide future visualizations of sensitive data.

As a co-creative method, collage can stimulate the design of collection visualizations by integrating diverse materials and perspectives. This retrospective reflects on a decade of workshops with over 15 partners in the arts and humanities, highlighting how this participatory format can bridge diverse backgrounds and generate insights and ideas.

This paper presents a comprehensive review of semantic annotation practices applied within the DH domain. Focusing on current methodologies, tools, and frameworks, we developed a multidimensional classification schema to assess annotation systems, along with a critical overview of semantic annotation in DH. The paper concludes with recommendations for future research.

This paper provides a conceptual overview of visualization-based storytelling tools developed in MEMORISE. We introduce a triangular definition of visualization-based storytelling, which we apply to the Heritage of Nazi Persecution (HNP). We introduce visitor-driven, expert-driven, and witness-driven storytelling, and we describe visualization and storytelling tools for diverse user groups.

This proposal aims to delineate the underlying concepts of the Cor framework’s adaptability and significance in the realm of digital humanities, emphasizing its diverse applications and innovative approaches.

“Race, Gender, and the Visual Culture of Domestic Labor” is a publicly accessible digital humanities project that presents an archive of tradecards and postcards depicting domestic labor, from the 1870s to the 1940s. In this paper, we describe the features of the archive and the techniques used to develop it.

Producing annotated corpora is essential for training annotation systems, but it is often a lengthy and expensive process. This paper introduces a method and a functional tool for transferring annotations from a source language to a target language, when relevant high-quality annotated corpora exist in a source language.

The digital publication of a vast and diverse cultural collection is the starting point for an investigation on the relationship between intermediality, narration, data and cultural heritage, that converge into the design of an interface model that allows visitors to experience textual and photographic archival items synchronously.

This presentation showcases the release of AREPR’s community archiving resource, Bilingual Archiving in a Box (BArch Box). Consisting of guides, manuals, and video tutorials on community archiving, BArch Box is a bilingual community archiving toolkit designed for use by community groups, universities, and libraries across the Spanish- and English-speaking world.

This paper examines how the 'anarchiving as research-creation' approach informs the Salvadoran Civil War Digital Music Archive. By blending historic and modern recordings, this digital repository explores music as a cultural memory and a tool for justice, addressing postwar challenges through experimental, participatory, and future-oriented archival practices.

This study leverages dynamic contextual embeddings to analyze conceptual evolution in 18th-century French texts. Employing fine-tuned BERT and CamemBERT models, we identify diachronic semantic shifts across historical subcorpora. Quantitative metrics and qualitative assessments reveal nuanced changes in key concepts land concept clusters, advancing methods in computational intellectual history.

This study presents a historical interpretation system using relationship network graphs to analyze power dynamics, exemplified by civil officials' military authority. By integrating Retrieval-Augmented Generation (RAG) with large language models (LLMs), the system retrieves and interprets relational data, uncovering hidden details and enhancing historical text analysis with clear, explainable outputs.

This paper discusses digital humanities approaches, including Named Entity Recognition, machine learning OCR methods, and topic modeling of a handwritten Indigenous Journal by John Norton, Teyoninhokarawen. This recently-discovered journal is an account of Norton's travels from Canada to America and Britain; we use DH method to analyze the journal.

We analyse collaboration networks in Mexican theatre productions from 1900 to 1989. Our results suggest that the periods with more stable funding tended to have more closely knit communities, and institutional eras are dominated by more stratified and distinct communities.

This study investigates regional variations in Japanese children’s songs (warabe uta) by classifying melodies into "word-based" and "melodic" types using machine learning and GIS tools. Results reveal distinctive regional and demographic trends, with Kyoto’s melodies more "melodic" and urban areas favoring "word-based" styles, highlighting sociocultural and environmental influences.

This study explores the collective memory and national identity of the Visegrad Group (Czech Republic, Hungary, Poland, Slovakia) using innovative methodologies such as digital humanities, network modeling, and audio spectral analysis. Combining historiographical analysis with emotional and spectral analyses, it reveals intergenerational and sociocultural dynamics shaping memory and identity.

Social scientists increasingly use NLP for large-scale text analysis but face programming challenges. CodeFlow automates code generation and optimization via LLMs, translating research goals into functional code. It achieved 0.95 accuracy in sentiment analysis with a BERT-based classifier, allowing researchers to focus on questions while ensuring computational rigor.

Pandore is a user-friendly toolkit for humanities and social sciences, enabling data collection, preparation, analysis, and visualization without advanced coding skills. Recent updates include bug fixes, interface enhancements, integration of modular Python scripts, a connection to Gallica, and deployment on a GPU-equipped server.

This paper is a presentation of Flow Filter - a generalisable exploratory visualisation tool and query builder designed to aid serendipitous discovery of large data sets and aid hypothesis formation. It will present the concept and rationale and illustrate its use and effectiveness via three case studies of historical datasets.

Open Science requires the development of specific skills, that can be named as Open Science Literacy (OSL), already described in a previous research. This new study intends to identify a set of elements that could fit the presented OSL scheme and propose a Digital Humanities OSL chart of competencies.

This study evaluates Large Language Models (LLMs) for Named Entity Recognition (NER) on a poetic form i.e., Marisya in the right to left Urdu script. The scarcity of annotated Urdu datasets by creating a human-annotated corpus is addressed and the performance of LLMs against the human-annotated corpus is evaluated.

In 2024, the UChicago Center for Digital Scholarship hosted its first Library and Emerging Technologies Summer Camp, a workshop series aimed at teaching the basics of Digital Scholarship and fostering opportunities for collaboration among library staff. This paper describes the lessons learned from this project and our hopes going forward.

The Edut 710 initiative addresses selection, attention, and dissemination biases in real-time digital archiving of mass atrocities, emphasizing accessibility. Using computational tools and iterative methods, it ensures inclusive representation of over 1,200 testimonies documenting the October 7, 2023 Hamas attack. This model redefines ethical, accessible digital archiving for contemporary events.

A pesquisa aborda os desafios éticos nas Humanidades Digitais no Brasil, explorando dilemas relacionados à privacidade, vieses algorítmicos e ciência aberta. Destaca a importância da ética reveladora (disclosive ethics) como ferramenta crítica para promover práticas responsáveis, justas e transparentes, visando fortalecer a integridade científica em um contexto de crescente complexidade tecnológica e informacional.

This paper advocates for an approach to DH pedagogy that integrates DH concepts with global cultural frameworks and narratives using historicization, contextualization, and analogizing as key moves. Combinatorial vector-based semantics is the proposal’s use-case concept, which is linked to inclusive, non-Western perspectives as illustrative of the approach.

We use computational text analysis and qualitative coding to explore how, when, and where “AI” and associated concepts/methods (like “LLM”) appear in course descriptions collected from the University of California and the California State University systems’ course catalogs for all departments, focusing on data for Academic Year 24-25.

FAIR principles guide best practice for research data management. FAIR means Findable, Accessible, Interoperable, and Reusable. To reuse digital objects, one needs to have full provenance information on the object and its metadata. This paper presents two approaches for tracking metadata changes. One uses Git, the other PID features.

How accurate is the work done by the volunteers who do most of the work on crowdsourcing projects in cultural heritage? We analyze the results of the 1970 MIssouri Death Certificate indexing project, applying traditional quality metrics from Optical Character Recignition and Handwritten Text Recognition to human-created text.

Aborda o uso do software Transkribus para transcrição de documentos históricos no ensino.

The paper discusses approaches to include retrocomputing and computer laboratories into Digital Humanities practice for research, didactics, preservation and self-reflection.

Le biblioteche moderne si configurano come spazi multifunzionali, dove fisico e digitale convergono per favorire creatività, confronto e inclusione. Attraverso servizi innovativi e collaborazioni strategiche, promuovono accessibilità e partecipazione, valorizzando diversità e competenze per abbattere barriere e stereotipi, con l'obiettivo di costruire una società equa e inclusiva.

Many provenance databases do not meet FAIR standards.This paper emphasises the need for FAIR provenance data and proposes a method to create structured and FAIR data that can be achieved by non-experts. It also critically discusses why FAIR provenance data may not always be better.

Advanced computational methods in digital humanities have increased demand for text-mining files, including third-party licensed datasets, which present data sharing challenges. This presentation explores navigating these challenges through a case study of a librarian assisting a PhD candidate in sharing licensed research data from various vendors.

The FAIR principles (Wilkinson et al. 2016) are important in sustainable research data management. Applying FAIR assessment tools in a real-world, domain-specific context, we find the overall FAIRness score and ranking roughly comparable between tools, while the individual categories (Findability, Accessibility, Interoperability, Reusablility) vary due to different test collections.

We suggest combining curation with research-driven approaches: (1) digitization and indexing of archival material as well as (2) collection and analysis of underlying meanings and perspectives of the actors through ethnographic methods. This is showcased regarding everyday culture surrounding community function halls with restaurants, central establishments for rural communities.

Presenters will discuss the evolution and future directions of the Kentucky Hispanic Heritage Project, highlighting how a backward design approach informed by community input has guided decision making about accessibility and source selection, positioning the project as engaged digital scholarship that celebrates the production of knowledge by the local community.

This study introduces Topic-Method Composition (TMC) to analyze the co-occurrence of research topics and methods in Digital Humanities. By constructing a TMC network from large-scale bibliographic data, it identifies key research paradigms, highlights DH’s interdisciplinary nature, and provides a replicable workflow for exploring topic-method relationships across academic disciplines.

Atualmente a preservação do património cultural carece de a inovação metodológica praticada para a gestão de coleções ser mais eficiente. Inovar metodologicamente a rastreabilidade através de métodos como QR Codes e RFID permitirá colmatar carências. Esta comunicação pretende explorar soluções aplicadas à museologia complementando o tradicional e a tecnologia.

This paper examines domain-specific knowledge graphs as transformative tools for Digital History, highlighting their ability to model complex relationships, support multilingual datasets, and integrate linked data essential for reconstructing fragmented narratives of sensitive events. It particularly explores Wikibase’s role in advancing historical research, cultural preservation, citizen participation, and open science.

This paper investigates the application of citizen science in cultural heritage through an ontology-based analysis of Scopus articles. By utilizing the Citizen Science Ontology, it examines project aims, tools, and outcomes. Findings underscore contributions, challenges, and future opportunities for advancing participatory approaches in preservation and promotion of cultural heritage.

Digitizing cultural heritage preserves artifacts and improves accessability. Libraries like the Jagiellonian Digital Library offer datasets via OAI-PMH, but incomplete metadata limits searchability. We propose using computer vision, AI, and semantic web technologies to enrich metadata and construct knowledge graphs for digitized manuscripts and incunabula.

This study explores how Generative AI and Retrieval Augmented Generation (RAG) enhance archival research by developing an AI-enhanced database for the Collection of Historical Archives on Sino-Russian Relations. Integrating metadata and thematic search capabilities, the methodology improves retrieval precision and accessibility, offering transformative potential for historical research across diverse domains.

This study introduces the process of creating a corpus of Ottoman Turkish poems written between 15th and 19th century and gives a use case for the corpus on the adaptation of the aruz meter in Ottoman Turkish poetry via using the corpus.

This paper examines Active Learning (AL) in vision models by asking: which data to train on, and how much? Using a case study on person detection in art-historical images, it discusses the potential of AL to improve model performance while providing broadly applicable insights for disciplines within image science.

As Indigenous peoples continue to advocate for their rights and wellbeing, including in the digital sphere, this paper outlines the recommendations of the iREAL project to support Research Technology Professionals and Librarians to undertake requirements elicitation for AI/Machine Learning projects in libraries incorporating Indigneous data in a respectful manner.

This paper introduces iberz, a bibliographic database of translations into Yiddish, and performs a quantiative analysis of its contents. The database contains 1,375 translations from 1868–1993, which are linked to source texts, This data is made publically available in a GitHub repository, as well as via a web app.

This presentation introduces a web-based platform designed to address the challenges of AI integration in writing instruction. The platform (https://aawe.ai) enables instructors to control AI assistance levels while providing students with a distraction-free writing environment. Preliminary classroom use yields promise in balancing technological support with academic integrity.

Esta apresentação pretende dar a conhecer a base de dados inclusiva Mind_the_Blind,os seus objetivos e impacto na comunidade portuguesa. Pretende também criar pontos de contacto com outros investigadores em países com necessidades semelhantes, promovendo a colaboração e o intercâmbio de melhores práticas na formação em acessibilidade aos meios de comunicação.

This research investigates the integration of Digital Humanities in translation studies education through a survey of postgraduate students in Hong Kong. The findings reveal a gap between students' recognition of DH's importance and their expertise. The study proposes strategic interventions for curriculum development to enhance DH competencies in translation pedagogy.

This short presentation focuses on a reproducible Information Retrieval Q&A pipeline tailored for historiographical corpora, specifically the theses abstracts of our university (3,000+ texts in French). It addresses retrieval challenges by integrating vector-based indexing, semantic search, and LLMs, offering structured, contextualized responses to enhance research and exploration in large corpora.

This study introduces SentiAnno, a sentiment-annotated, topic-specific corpus of Austrian historical newspapers (1700–1938). Focusing on the topics of migration and minorities, SentiAnno enables fine-tuning of LLMs for sentiment analysis and topic classification. Annotation processes, tools, and inter-annotator agreement are described, with the final corpus to be published on Zenodo, supporting FAIR principles.

FLAME, using Leave-N-Out grams, detects formulaic language in medieval charters despite variations in wording and structure. It overcomes limitations of traditional n-gram and skip-gram approaches by flexibly capturing long-range dependencies and identifying functional equivalence across diverse expressions. FLAME facilitates analysis of formulaic language evolution, revealing flexible patterns in legal language.

We use BERTopic to analyse themes in the bilingual speeches of the Carniolan Provincial Assembly. We examine common topics discussed in the sessions and how they change over time to gain insight into the key societal issues at the regional level in the Austro-Hungarian Empire at the turn of the 19th century.

Wie verändern digitale Ausstellungen unsere Wahrnehmung und Interaktion mit kulturellem Erbe? Das Forschungsprojekt an der TU Chemnitz untersucht diese Frage durch eine innovative Kombination aus Korpusanalyse, multimodaler Annotationsmethodik und experimentellen Studien. Ziel ist es, empirische Erkenntnisse zu Design und Nutzererfahrung zu gewinnen und praxisnahe Handlungsempfehlungen für die Kulturvermittlung zu entwickeln.

UniTermGPT explores ChatGPT’s handling of German higher education terminology across Austrian, German and South Tyrolean varieties. By compiling a specialized corpus, applying prompt engineering and evaluating translations, it addresses language-variety-specific terminology challenges in LLMs. The project highlights the societal relevance of terminology, offering open research data and practical recommendations.

The increasing use of data-driven research in the field of digital humanities has emphasized the fundamental importance of research data management (RDM) and data stewardship skills. This contribution highlights the importance of education in these areas to advance open research data, uphold the FAIR principles, and promote sound scientific practices.

The project explores museum didactics with a focus on visual education, using photography and innovative technologies in order to promote experiential and inclusive learning beyond the classroom that integrates the relationships between schools, museums and the territory.

This poster describes ongoing collaborative digital research on the Bulletin of the Center for Children’s Books, a children’s literature review journal founded in 1947 that provides a vital record of the history of children’s book publishing and professional children’s book reviewing during the 20th and 21st centuries.

“Putting WKZO on the Map” uses mapping and encoding to solve problems facing audiovisual collections. Recognizing that local radio exists in a local landscape, this project maps a radio program through locating companies featured and encoding transcripts to identify the names of people and places mentioned.

This study proposes a Dual-Track Activation Model of Documentary Heritage based on LLMs. The model addresses the challenges of utilizing specialized, multimodal heritage resources. It is validated through the development of a knowledge base platform for the Suzhou Silk Archives as a case study.

Small grants play a crucial role in driving innovation and inclusivity in Digital Humanities by supporting interdisciplinary research, data analysis, and Open Science. With fewer bureaucratic barriers, they enable early-career and independent researchers to experiment, collaborate, and create open-access resources, fostering rapid methodological advancements and broadening academic participation.

This poster describes how the COMMONS project, involving three French research infrastructures, aims to address the needs related to the creation and use of links between data and publications: from technical requirements to creation of complex publications (ie. data papers, data displayed in an article etc.)

Handwritten codices have been systematically cataloged for centuries. Today, hundreds of thousands of catalog entries can be accessed digitally. As a proof of concept, I scrutinized the current possibilities in accessing, harvesting, curating, and processing this domain of knowledge to create a visual tool for further analysis and heuristic research.

The Grounded AI Map visualizes AI’s involvement in science through an “irreductionist” lens. It supports exploratory hermeneutics through computational annotation and physicalization, engaging audiences interactively while preserving data ambiguity, polyvalence and contradiction.

This study explores how citizen science can contribute to enhancing the readability of hidden historical text on book bindings or palimpsests,

This poster discusses the challenges of curating, preserving, and ensuring access to internet content, specifically within the context of Polish digital culture. By examining the ongoing iPBL project, the research highlights the complexities involved in selecting, archiving, and sharing ephemeral online resources.

O estudo pretende apresentar resultados preliminares de uma scoping review sobre a adoção de práticas de Investigação Aberta nas Humanidades Digitais.

Text reuse has been used to trace the flow and diffusion of policy ideas in legislatures through bill-to-bill and model-legislation-to-bill comparisons. This study investigates how ideas and rhetorical strategies refined in private communications between anti-trans political actors have influenced bills proposed in the United States between 2019-2024.

My PhD project is focused on creating a computational model to predict the auction prices of manuscripts and early books. This abstract summarizes my current progress with the challenge of extracting large amounts of data from auction catalogue texts, and testing the performance of GPT4 as an annotation assistant.

Amid widespread decline, tracking the UK local media sector is challenging due to outdated directories and rapid changes. To address this, we developed a semi-automated system using OSINT to monitor closures, launches, and ownership changes. This model enhances accuracy, reduces labor, and informs policy on media pluralism and sustainability.

The use of digital methods and tools is an integral part of humanities research. Smaller humanities disciplines run the risk of not keeping pace with the digital transformation. Using the example of Byzantine Studies, we want to discuss how small disciplines can be strengthened in the face of digital change.

This poster familiarizes digital humanists with:

1. what zines are

2. what zines can make possible for the digital humanities

The Zine Bakery project is a portal into zines as a welcoming, inexpensive, effective format for do-it-yourself DH scholarly communication and public outreach; friendly digital method teaching; and zine-inspired DH research explorations.

Based on sociological survey data on the cultural participation of German citizens, this poster outlines an early-stage PhD project aiming to develop data-driven profiles of cultural participation behavior to characterize readers and non-readers by their leisure activities. Factor analysis and cluster analysis serve as methods to establish the profiles.

This poster presents a RAG system prototype developed within the AHRC-funded iREAL project, exploring ethical AI implementation with Indigenous cultural data. Built using open-source models and technologies, the system demonstrates how open-source tools can responsibly process sensitive cultural materials while maintaining transparency through hybrid search and observability features.

We explore enhancing LLMs' ability to generate personality-consistent character utterances for novels. We annotated the personality traits of characters from 233 novels and observed that characters with similar personalities exhibit similar linguistic patterns in their utterances. Llama-2-7B was trained on character utterances using instruction tuning, producing more personality-consistent utterances.

This study explores gender representation in original Hungarian children’s literature and its translations into English, German and Dutch (1925–2025) using computational linguistics methods. Early findings highlight linguistic patterns in pronouns, adjectives, and occupational titles, revealing shifts influenced by sociocultural changes. The poster presents preliminary insights from corpus development and analysis.

This study examines Domenico Gerosolimitano's 17th-century Hebrew translation of the New Testament using DICTA's digital tools. A Jewish convert to Christianity, Domenico's work offers unique insights into early modern religious translation practices. Despite claiming multiple source texts, preliminary findings suggest his translation primarily follows the Peshitta version, reflecting complex cultural and theological negotiations.

The School of Salamanca. A digital collection of sources offers central works by Salmantine authors. The print edition Las Siete Partidas (1555) is a complex dual text structure (Spanish main text/Latin gloss) that is demanding for TEI-transcription and digital representation. The poster presents our workflow to successfully address these challenges.

We focus on an exploration of features for developing metrics to quantify the degree of changes in digital online projects over time. Our purpose is to provide systematic methods to sustain and preserve culture and the digital scholarly infrastructure in the humanities over time, preventing their degradation and decay.

The Citizen Archive Platform simplifies the preservation of born-digital cultural heritage by enabling citizens to submit data seamlessly to institutions like archives and museums. Developed under the "Dialog City" initiative, the CAP standardises data transfer, ensuring accessibility, usability, and integration into OAIS while addressing key challenges in digital preservation.

CorpSum is a web application that enables user-friendly, dynamically generated queries in text corpora along different extralinguistic extralinguistic dimensions of variation (such as the dimensions time and space). It is a bespoke software module originally developed at the ACDH-CH to facilitate work with the Austrian Media Corpus (AMC)

The HAICu project leverages artificial intelligence and advanced machine learning to transform digital humanities research. By analyzing handwritten manuscripts from Dutch archives, researchers develop innovative computational techniques that cluster document layouts, generate metadata, and create new pathways for understanding historical collections through a collaborative, human-in-the-loop approach.

The poster presents an innovative interdisciplinary research hub, dedicated to digitality, outlining the centre's structure and collaborative approach. The focus lies on understanding digitality's epistemological nature and its potential to contribute to DH. The Reading and Writing Lab's work is highlighted, aiming to provide a theoretical framework for digitality.

This contribution presents the ongoing development of the Voci dall'Inferno project. This research initiative aims to create a digital corpus of non-literary testimonies from Lager survivors and analyze it to identify expressions from Dante's Commedia that witnesses use to describe their harrowing experiences.

VERA is a digital collaboratory for participatory research in the social sciences and humanities. Developed through the COESO project and now part of OPERAS, it enables multilingual collaboration between researchers and citizen scientists, fostering inclusivity, knowledge exchange, and innovative research practices across Europe.

EU Horizon2020-Funded Computational Literary Studies Infrastructure (CLS INFRA) is reaching the end of a four-year journey towards shared and sustainable infrastructures within the FAIR and CARE principles. This poster presents the outputs of the CLS INFRA project 2024-2025, focusing on the resources that open multilingual, participatory digital practices to all.

The software offers two retrieval functions. Firstly, it enables users to retrieve the small seal script characters that serve as basic elements through the number of strokes and stroke sequences. Secondly, it allows users to retrieve other small seal script characters composed of basic elements by means of IDS.

Ruiju Myōgishō (11th century) extensively cites Buddhist scriptures and classical texts with detailed source annotations. Combining TEI/XML and RDF effectively models its intricate structure, especially source data. This poster presents a model for encoding source information, highlights technical challenges, and explores its implications for early Japanese dictionaries.

O GIMTE, vinculado ao MATLIT LAB, explora a codificação textual com XML-TEI no ensino. Esta proposta de póster apresenta o trabalho realizado pelo grupo de discentes na transcrição, marcação semântica e tradução para o romeno de textos de Fernando Pessoa, no contexto de experimentação do laboratório.

In our project on Buddhist murals of Kucha, Xinjiang, we developed an interactive system for visualizing their locations. Using SVG-based register systems and 3D-cave models, it enables spatial analysis and accessibility. This approach improves understanding of mural arrangements, with potential applications in digital humanities for analyzing complex artworks.

The digital platform “Siberiana” (siberiana.online) for cultural heritage collection, preservation, and actualization is developed at the Siberian Federal University (Krasnoyarsk) as part of the Institute of Digital Humanitarian Research project.

This research explores 19th-century American serial fiction through data from Chronicling America, using computational methods to map networks of serialization. By analyzing formal features like chapter headings and author names, I uncover patterns of publication, reprinting, and reader engagement, recovering forgotten authors and rethinking seriality's role in literary history.

The eArchiving initiative provides guidance for digital preservation. The eArchiving Reference Curriculum is a master's level framework covering key aspects for that purpose, including data integrity, security, and long-term accessibility, intending to be a guide for academics and students. This poster will present the core elements of this framework.

The Urban Video Archive (UVA) is a digital repository documenting video activism in Rio de Janeiro (2013–2023). Developed with Brazilian media activists, it emphasizes co-creation and community archiving over institutional archiving and social media sensationalism. It highlights marginalized communities’ urban struggles through an interactive map, networked videos, and open-access tools.

This poster presents the Digital Camerarius, a digital edition of the Symbola et Emblemata by Joachim Camerarius. The goal of the project is to provide a machine-readable transcription enriched with structural and semantic markup, and to facilitate multimodal exploration with Retrieval Augmented Generation (RAG).

This poster will showcase our work-in-progress digital edition of John Derricke’s The Image of Irelande (1581), focusing specifically on how the PIs have worked with its visual elements. This poster presentation demonstrates how TEI can offer opportunities to enhance textuality and storytelling through access and accessibility.

This panel examines the ethics and emergent challenges of what "open" now means in the current age of AI. The four papers each engage with this question from different though related perspectives: data sovreignty, project design and privacy, pedagogy, and artistic labor.

This panel showcases the need for stronger collaboration between research infrastructures enabling data FAIR-ness, training programmes ensuring competent reuse of language data and data journals establishing rigorous review processes. This is essential to ensure data quality, relevance, and impact, maximising its potential for reuse in research, education and societal contexts.

This study compares humans and AI in recreating a three-panel Nancy cartoon. Humans, with basic art training, excelled in creating coherent visual narratives, while AI, despite impressive artistic replication, struggled with storytelling. The results highlight human superiority in transforming instructions into meaningful stories.

This paper introduces a multidimensional similarity search tool, Motif-Match, for digital art history, emphasizing similarity as multi-dimensional, cumulative, and situational. Through participatory design and user evaluation with 39 participants, we explore the roles of control and transparency, offering insights into balancing technical innovation with the nuanced needs of humanities research.

This presentation explores the challenges and opportunities of implementing open data in Catalan archaeology. It examines the current infrastructure, researchers' practices, and barriers to data openness. The study provides recommendations to promote a new research culture, with the goal to lead a smooth transition to open archaeological research.

I build on my previous research on "time maps" by expanding their theory and demonstrating their interpretive utility. Time maps are the graphs which are produced when a narrative’s fabula is plotted against its syuzhet. I introduce three advanced theoretical concepts, then use time maps to close-read several narratives.

This study examines subset selection in bibliographic research, focusing on Finnish literary history (1809–1917). Comparing manual and automated curation, we highlight their respective strengths and limitations. We propose a hybrid approach combining automation for scalability and manual curation for precision. Our findings enhance transparency, accuracy, and reproducibility in literary datasets.

A comparative study of the performance of large language models vs. metric analysis of Spanish folk song lyrics existing datasets, with qualitative data.

Poetic meters impose recurrent patterns on a language already dense with structured relationships. In large corpora, metrical effects accumulate into strong statistical regularities, becoming a major source of linguistic variation. In this paper we demonstrate how different meters distinctly shape seemingly unrelated feature distributions in Czech, German, and Russian corpora.

Esta proposta apresenta o Projeto Joanina Digital, dedicado à digitalização de cerca de 30 mil volumes da Biblioteca Joanina e à criação de uma plataforma multifuncional. Com base na teoria da remediação, serão explorados os desafios e os impactos técnicos, culturais e investigativos do projeto.

Heritage of Nazi persecution (HNP) poses a challenge for computer-based visualisation and design, which gives rise to ethical considerations. In this paper, we discuss principles for digital reconstruction and accessibility of historical sources, and relate this to solutions developed for an inclusive design and visualization of HNP.

O artigo aborda a diversidade linguística nas Humanidades Digitais (HD), com análise bibliométrica das bases Web of Science e Scopus (2012–2021). Apesar do predomínio do inglês, observa-se crescente multilinguismo, com o espanhol e o alemão em destaque. Os resultados sublinham a importância da inclusão linguística nas HD.

This paper critiques "data colonialism" as a metaphor for exploitation in the digital economy, arguing it misses key aspects of contemporary data practices. We propose instead the "company town," which better captures the user-platform relationship and highlights ethical concerns for Digital Humanities researchers involved in community-focused work.

This paper addresses issues of sustainability of digital resources, their use and reuse, particularly from the perspective of research infrastructures. We argue that research infrastructures – through the combined efforts of conceptual rethinking, technological solutions and strategic advocacy – have the potential to transform how we sustain and engage with DH scholarship.

This study evaluates five unsupervised sentiment analysis methods on Early Modern German and English texts, addressing challenges like semantic shifts and limited resources. Findings reveal significant limitations in current approaches, emphasizing the need for domain-specific-models, multilingual resources, and hybrid methodologies to enhance sentiment analysis for historical datasets and heritage preservation.

Il contributo che proponiamo è indirizzato a presentare l'avanzamento di un progetto di ricerca volto a definire un modello di digitalizzazione delle opere custodite nei depositi in seguito agli eventi sismici che hanno colpito l’Italia centrale tra il 2009 e il 2016.

This presentation explores the paradox of increased access to digital tools for documentation and archiving. While access empowers community-driven efforts, it also exacerbates challenges such as market saturation, unpaid labor, institutional dependency, misinformation, and external manipulation. Case studies from Brazil and Iran reveal how accessibility can undermine autonomy and accuracy.

This paper explores a crisis in global AI Art culture, which tends to be appropriated by corporate entities that do not respect the principles of FAIR and CARE. What this means is that AI, despite its technological potential, more likely exacerbates inequality and discrimination in collectives like indigenous cultural expressions.

The Journal of Open Humanities Data supports FAIR data sharing and reuse through peer-reviewed articles. In 2024, a special collection of papers titled Amplifying GLAM Collections: Scalable and Inclusive Data Practices was created to increase representation of cultural heritage datasets and practices. This paper will describe the results and implications.

Este panel reúne proyectos que han construido infraestructura digital colaborativa y abierta para la preservación, análisis y acceso a la documentación en contextos de bajos recursos, incluyendo infraestructuras para digitalizar, sistematizar, interrelacionar, analizar documentación de archivo y desarrollar nuevas estrategias pedagógicas y de divulgación del conocimiento histórico.

This panel explores diverse dimensions of openness within the galleries, libraries, archives and museums (GLAM) sector globally, shaping discussions about accessibility, inclusivity, participation, and knowledge democratisation. Cultural heritage institutions are responsible “to all citizens”. Yet there are gaps relating to collections, knowledge, policy, technology, engagement, IP, ethics, infrastructure and AI.

Focusing on predictive algorithms and AI technologies in law enforcement, this paper argues that a digital humanist inquiry of the historical development of law enforcement in the United States is necessary for identifying how emerging policing technologies perpetuates systemic harm against marginalized communities by design.

We investigate the effectiveness of LLMs in evaluating text similarity, a fundamental task in CH research. We study the similarity of German poems from different perspectives, such as content or form. Our results show that recent commercial models are comparable to or better than supervised models (zeroshot, chain of thought).

This study employs natural language processing and machine learning to analyze religious trajectories in Holocaust survivor testimonies. Utilizing large language models, we reveal patterns of evolution in beliefs and practices under extreme conditions, offering insights into thematic narrative development and demonstrating the transformative potential of computational methods in historical analysis.

This study employs a computational approach to analyze game mechanics using a collection of 296 predefined mechanics. By identifying their occurrences across a large dataset of games, we reveal trends in their popularity, evolution over time, and their relationships to game genres, demonstrating the method's potential for "computational game studies".

In this bi-national study comparing attitudes towards digital humanities in China and the UK, we explore interviewee responses towards a number of questions around DH identity formation, research infrastructures and professional structures. We discuss proposals to foster greater transnational exchange, using China and the UK as a case study.

Digital Humanities must critically study datasets to avoid intrinsic bias. This paper analyses bias in 13 datasets from the largest meta-collection of hate speech datasets, discovering hidden bias as a temporal trend of reduced lexical variability and dispersion, and a disproportionate focus on specific social groups or language types.

This methodological contribution deals with the problem of operationalising dramatic texts. More specifically, it introduces vectorisation according to structural features as a relatively novel and efficient option for accomplishing this task. Furthermore, it discusses potential and limitations of this methodology and presents some of its most recent applications in research.

This study presents a corpus-based approach to creating a SKOS vocabulary tailored for Ukrainian epigraphy. Integrating digital tools, NLP, and FAIR principles addresses gaps in cultural heritage preservation, offering scalable, efficient methods to document, analyze, and promote Ukrainian inscriptions while ensuring global research interoperability.

Set against the backdrop of recent debates in postcolonial studies, this paper uses a series of quantitative proxies for the categories of "literariness" and "cosmopolitanism" to situate the works of the postcolonial authors writing in English within a larger corpus of translations from South Asian and European languages.

O estudo analisa a representatividade de género nas Humanidades Digitais, examinando padrões de autoria na Digital Humanities Quarterly entre 2015-2024. Os resultados preliminares revelam 50,9% de autoras no período compreendido entre 2018-2022, desafiando pressupostos anteriores sobre disparidades de género na publicação científica naquela área.

We apply distant-viewing analyses of pose and action recognition data to 30 full-length recorded works from three prominent theater directors (10 per director) to explore how computational methods can detect a director’s oeuvre-scale choreographic tendencies from video sources. We further evaluate which features best delineate such stylistic “signatures.”

This study leverages a LLM-based wordplay detection pipeline for authorship attribution in Colette's disputed works. Combining semantic segmentation, emotion filtering, named entity recognition and wordplay annotation, we detect a few intricate wordplays consistent with Willy's style. Results support minimal direct influence from Willy while identifying targeted passages, offering insights into collaborative authorship processes.

How can we improve quantitative analysis by treating absence not as a lack of information, but as a different type of data? Using a seemingly complete dataset as a case study, we draw on postcolonial theory, intangible cultural heritage, and anthropology to explore what absence conveys about DH practices.

La ponencia se centra en recuperar el legado de María Lejárraga en el Teatro de Arte, analizando su papel en la incorporación de danza y música. Mediante análisis de redes basado en correspondencia, memorias y programas de mano, se demuestra cómo sus colaboraciones transformaron piezas teatrales en producciones coreográficas innovadoras.

Many African policies position both digital /AI technologies and cultural industries as engines of sustainable development, but use of AI can also undermine artist livelihoods. The paper will consider how African policy frameworks could support communities in managing, protecting and promoting their digital cultural information in the age of AI.

Cette étude présente la création d’un corpus de romans du 19ᵉ siècle annotés en entités nommées dans leur intégralité, et l’élaboration d’un modèle de reconnaissance d’entités nommées adapté à de tels longs textes littéraires, et disponible librement en ligne. Nous évaluons ses performances, démontrant sa précision et sa robustesse.

This paper investigates the problem of synonymy in the langugage, namely is a classifier based only on word embeddingsl able to correctly classify synonyms according to their origin.

Although the language used for this analysis is Modern Hindi—significantly underrepresented in contemporary language research—the methodology presented is language-agnostic.