This research examines a critical gap in current digital text infrastructure for individuals with print disabilities: the systematic inaccessibility of specialized Unicode-encoded characters essential for academic discourse. Our investigation showcases the development and implementation of Niv Louie, an innovative software solution designed to facilitate access to previously inaccessible digital content.

The presentation introduces a Writing Assistant System designed to simplify and enhance the creation of easy and plain language texts in Italian. Addressing communication barriers in public administration and supporting inclusivity, WAS provides real-time feedback and educational tools, combining AI-driven suggestions with human oversight to improve accessibility and writing skills.

This presentation examines critical digital literacy as a multifaceted competency. Drawing on a Master’s-level course where students create 3D scholarly editions of toys within an authentic learning environment, we demonstrate how they develop skills, including critical and creative making and collaborative problem-solving –transcending discipline-specific knowledge to prepare for the digital and creative economy.

The panel discussion addresses the need for a new data culture in history and beyond with the aims to understand the fundamental epistemological affordances of the post-digital moment; to develop the necessary quotidian practices and disciplinary protocols; and to negotiate new understandings of history as a discipline of societal relevance.

This short paper asks: what can be discovered about the poetess tradition of nineteenth-century Britain–a mode of writing initially lost to literary history and recovered by feminist literary scholars–when using topic modelling to conduct a distant reading of primary, critical, and bio-critical materials?

Picture book covers carry abundant visual information about the story. They also indicate gender information through titles and illustrations, which might further impact preschool children’s gender perception. This paper will investigate how ChatGPT infer gender on covers of 6,629 preschool children’s books from the 2012-2022 National Library of China catalogue.

This presentation will introduce the work conducted by “The Adaptive Text Encoding Initiative Network,” an interdisciplinary research group of PhD students, faculty, and DH librarians in Canada. We will report on our progress as a team and provide insights into our proposed antiracist, decolonial, and inclusive TEI guidelines.

This study examines gendered communication in German-speaking Twitch gaming chats using BERTopic and qualitative analysis. Male streamers’ chats focus on gaming, while female streamers evoke social and emotional messages. No evidence of objectification was found, highlighting shifts in behavior or effective moderation. Findings underscore gender’s role in shaping digital interactions.

This talk explores how documentation can serve as a powerful tool for positive change by making our research and datasets more transparent. Detailed documentation not only facilitates the effective and responsible reuse of datasets and algorithms but also promotes more inclusive scholarship and ethical research outcomes.

Esta apresentação se concentrará no trabalho realizado durante o projeto de colaboração Text+ "Pessoa Digital", discutindo a importância das mudanças implementadas para garantir a sustentabilidade e modernizar a infraestrutura digital da edição acadêmica digital “Fernando Pessoa. Projects and Publications” e como elas se encaixam na infraestrutura do consórcio Text+.

Se analiza a las Bibliotecas Públicas de Bogotá como plataformas de co-creación de narrativas digitales de historia pública entre 2018 y 2021. A través del análisis de Los bogotanos del Bogotazo y Marca de agua se destaca cómo la mediación digital desafía narrativas oficiales y democratiza el conocimiento histórico.

Este artigo apresenta uma revisão crítica da literatura dos princípios éticos no universo da Inteligência Artificial (IA), investigando a sua conexão com as Humanidades Digitais (HD). Tem como objetivo refletir sobre o campo epistémico e identificar os desafios, práticas e recomendações, através de uma abordagem qualitativa.

Small-to-medium galleries face significant challenges in preserving and sharing their contributions, relying on spreadsheets or paper records.This issue is exacerbated by proprietary data platforms, which reduce data sovereignty. This paper presents a low-tech solution developed to normalize, preserve, and share gallery data, emphasizing sustainability, inclusivity, and resistance to digital monopolies.

With the project ‘Collation of Multilingual Text’ (COMUTE), we tackle the problem of aligning multilingual versions of text that are not literal translations, but which have undergone a major revision process including extensive additions, deletions and rephrasings. Our approach based on a hierarchical alignment solves issues of current state-of-the-art methods.

Der Beitrag widmet sich der digitalen Modellierung von Dekanonisierungsprozessen serieller Frauenliteratur im 19. Jahrhundert und ihrer theoretischen Reflexion. Hierzu untersuchen wir Relationen zwischen Texten in Frauenzeitungen und historischen Kontexten. In multivariaten NLP-basierten Analysen werden Grade der vielfältigen Einflussfaktoren gemessen und Bedingungen der Evolution serieller Literatur in deutschsprachigen Frauenzeitschriften miteinander verglichen.

Glossaries appended to monographic treatments of historical mathematical Chinese texts are one of the most important references available for technical terms in that type of language. In this study, digital methods are used to explore such glossaries, with an focus on understanding what phenomena from the original texts are included.

I will discuss the challenges of handling living resources and relate them to already existing strategies and their vulnerabilities. I will present an approach for managing living resources by considering the responsibilities of different stakeholders – researchers, funding institutions and data centers/libraries – and argue for their orchestration.

We describes workflows, complexities and challenges in a digital pilot project on the Archive of German Speaking Jews in Israel, concentrating specifically on two complementing classification workflows: image classification of the digitized scans as a preliminary step for document analysis, and LLM assisted parsing and classification of folder level catalog entries.

In this long paper a visual interface, the "ReFa Reader", which combines narrations and the exploration of semantic graph structures is presented. After a brief introduction of the functionalities of the Reader, three workflows on how to reuse this Open Source Prototype by any project with semantic data are elaborated.

On a dataset containing correspondences from the literary period of Early Romanticism, this submission aims to uncover gaps and biases in the data, investigate their impact, and mitigate their effect. For this purpose, it develops a mixed methods-approach including the use of knowledge graphs, data visualisation, and historical network research.

The Linked Infrastructure for Networked Cultural Scholarship (LINCS) mobilizes Canadian scholarship through making, disseminating, and promoting the use of linked open data (LOD). This paper will reflect on the user-experience (UX) work done at LINCS, focussing on the roles of context and agency in supporting serendipity in a linked-data environment.

Our project introduces a BERT-based machine learning model of generic mixture for classifying individual passages from nineteenth-century Gothic novels by their relationship to other, non-Gothic, contemporaneous genres. Our method offers not only important insights into individual novels, but also a new way to track generic transformation over time.

This paper uses plagiarism detection software and social network analysis methodologies to compare nineteenth-century American and Hebrew journalistic networks. Results reveal stark differences in frequency, context, and style of textual reuse. The differences reflect distinct cultural roles of the media in shaping public knowledge and identity.

Este trabalho traz uma reflexão crítica, embasada em um estudo de caso, sobre a tendência dos métodos computacionais de análise textual priorizarem o que estatisticamente se repete, dificultando o acesso do analista do discurso a certas nuances. Argumento pelo uso de métodos mistos para acessar o 'não-dito' pelos jornais.

This study explores how the American magazine The Delineator reflects the evolving representation of the 'New Woman' from 1894 to 1914 through its visual content. By employing artificial intelligence techniques, particularly CLIP, the research shifts focus from text to visual analysis, revealing insights into societal perceptions of femininity.

This paper demonstrates how ChatGPT simplifies SKOS thesauri creation from hand-drawn sketches or digital drafts, improving efficiency over traditional editors. Testing with DH and fictional taxonomies reveals high accuracy but minor errors. While less suited for large thesauri, this method promotes FAIR data practices and facilitates SKOS thesauri development.

This paper introduces an interdisciplinary pilot project centered on the automatic transcription of Catalan manuscripts from the Late Middle Ages, focusing on notarial documentation. We describe the creation of a new dataset for our initial experiments. The resulting datasets, models, and code will be made publicly available.

We describe advances and results in developing four accurate Handwritten Text Recognition models for the automatic transcription of Itálica cursiva, Procesal simple, Redonda, and Procesal encadenada calligraphies, the most prevalent in 16^th-17^th-century Spanish American historical documents.

This study explores the use of large language models (LLMs) for correcting OCR errors in 19th-century French texts. Despite its advanced capabilities, fine-tuned models faced challenges with generalization, increasing error rates. The findings highlight limitations of LLMs in character-level OCR corrections and point to future research directions.

The AIAI Data Collective applies Critical AI frameworks to latent questions of AI data storage. We are developing a digital tool that guides users through a decision-making process for ethical AI data management, resulting in recommended practices that address critical and ethical concerns such as labor, privacy, and bias.

This research proposal explores the intersection of Critical Digital Humanities and Generative AI, aiming to enhance critical thinking in education. Through a mixed-methods approach, it will develop practical guidelines for educators, addressing challenges such as bias and transparency while fostering thoughtful engagement with AI technologies.

This paper, though examining discourses on access to digital archives (particularly community digital archives) and discussion on building care in access from the Stories on Contested Histories International Programme 2024, argues that studying challenges of access to digital archives and collections can help in conceptualising frameworks for inclusive access.

The paper invokes three narratives of disruptive digital projects that make comics accessible through AltNarrative, a multilingual digital repository, or the inception of a lab to create a born-digital and an inclusive comic of the future. The issue of web or digital accessibility is evaluating with respect to WCAG2.2 compliance.

This paper presents a framework for integrating LLMs into critical research work flows, addressing legal, ethical, and methodological challenges. Drawing on projects analyzing historical newspapers and court records, it emphasizes aligning LLM use with established hermeneutical practices to navigate automation responsibly and set standards for DH research and beyond.

The project marks the second phase of digitizing Tolstoy’s legacy, expanding beyond the 90-volume collection. It focuses on seven sources, each reflecting the writer’s life and works, connected by common temporal points. We discuss the preparation, markup, and presentation of over 32,000 documents through an interactive interface.

The project Auden in Austria Digital explores novel methods of formalizing interpretation in scholarly digital editions by embedding RDFa within TEI/XML. By transforming editorial argumentation into machine-readable data, the project seeks to make interpretation, interpretational responsibility, and scholarly uncertainty accessible to computational processing.

This project reevaluates the historiographical practices of Dai Nihon Shiryo, focusing on the functional representation of Bouchū (marginal notes) and Warichū (interlinear annotations) using TEI. By addressing their embedded temporal semantics and editorial distinctions, this study uncovers implicit historical perspectives and contributes to advancing TEI applications in digital scholarly editing.

This paper presents the open-access semantic edition-archive of Cesare Pavese’s literary and documentary heritage, developed within the PAVES-e project. It employs the CHROMA model, incorporating XML-TEI encoding, ontology-driven semantic organization, and interactive visualization tools to facilitate enhanced accessibility and semantic and lexicographic analysis of Pavese’s works.

This paper focuses on the modeling of the digital critical edition of a collection of Aesopic fables entitled Isopet 1-Avionnet, dating from the late 13th and early 14th centuries. It will notably involve translating, into the XML/TEI encoding of this collection, the textual, rhetorical, narrative and intertextual structures of this text.

We present a method for inferring semantic social networks from textual data by analyzing terminological similarities using topic modeling. Applied to a corpus of 3,698 scientific articles in astrobiology, this approach identifies "hidden communities of interest" (HCoIs)—groups with shared semantic content—and enables diachronic analysis of community evolution.

This study employs Network Theory to investigates the similarities and differences in co-occurrences within Ancient Greek and Latin texts from pre-Christianity and during the early Christian period. The case study focusses on mentions surrounding mentions of Plato in these texts.

This paper examines debates on a German anti-hate speech law called Netzwerkdurchsetzungsgesetz (NetzDG) on Twitter/X. By applying network analysis, NLP methods, and close reading, it investigates content published by the most retweeted accounts mentioning the law, considering their potentially high influence in shaping the discussion on the platform.

This work discusses ongoing tests and development of an R Shiny Web Application to visualize a dataset of National Health Service, Centers for Disease Control and Prevention, and World Health Organization Tweets during 2020. Interactive visualizations display different text analytics using tidytext and quanteda.

Gephi Lite, a web based and lighter version of Gephi, aims at pursuing Gephi original ambition: democratizing network visualisation edition through visual means. In this paper we present what the web does to network visualisation edition while presenting Gephi Lite main features.

The presentation showcases first results of a project aimed at representing the editorial transformations of the Children's and Household Tales of the Grimm brothers and demonstrates the variety of changes on several selected fairy tales. It further proposes a first insight into a scalable framework to visualize the editorial changes.

In 2024 I coauthored a forthcoming book, The Routledge Guide to Digital Humanities and Canadian Literature. This book presented a number of insights and challegns relevant to anyone attempting to write about the history and practice digital huamnities for a non-specialist audience. In this paper I will present lessons learned.

This presentation employs computational stylistics to analyze Yi Sang and Yŏm Sang-sŏp's works from Korea's colonial period. Using a custom sensory classification model, it quantifies stylistic and aesthetic differences with a focus on the body, revealing broader affective politics through characters' physical and psychological responses, extending beyond conventional emotion analysis.

This paper examined the relationship between the social situation of war and war writings in the British Romantic period by reviewing the topic modeling results and examining the visualized changes in war topics in books published between 1740 and 1840.

Presentación de la etapa inicial del proyecto de investigación Trad-Divers, consistente en la construcción de un dataset, en acceso abierto, modelando datos de 8 programas de ayuda a la traducción editorial en Iberoamérica (2001-2022), con el objetivo de incidir en la discusión y diseño de este tipo de políticas culturales.

This research uses Omeka S to construct and analyze a parallel corpus of 19th-early 20th century Japanese and Ryukyuan Bible translations. Using TEI encoding, stylometric analysis, and LOD principles, it reveals translation networks and linguistic patterns among Protestant, Orthodox, Catholic, and regional language translations, preserving linguistic heritage through digital methods.

We proposes a new method for extracting text in speech bubbles, onomatopoeia, and annotations from manga by identifying differential regions in multiple language versions. A preliminary experiment demonstrates the effectiveness of our method, achieving high accuracy in extracting speech bubbles and other areas.

This project develops a context-sensitive parser for Syriac using an encoder-only transformer model within ETCBC’s Qoroyo platform, addressing Semitic text challenges. It benchmarks LLMs for ancient languages, aiming for high accuracy and a quantifiable morphological analysis evaluation like GLUE or BLEU.

The study aims to test whether the well known stylometric methods are able to distinguish machine and human literary translation. The analysed corpus consists of novels translated into English, French and Polish to test whether the inflectionality of language plays a role in the quality of translation.

Previous work showed major differences in social networks of main characters and their relations of translations of a narrative. Here, we investigate reasons why this is the case. Quality of named entity recognition has the largest impact, while differences in language preferences do not have a major impact.

This submission will focus on a recent locative narrative in the context of contemporary Greek literary production (2021-2022)- Ismini Gatos' america² -and will reveal that space, time and body, in physical or digital terms can recontextualize the relationship of any user with the local environment and also with themselves.

This study constructs a digital narrative model for Chinese historical classics, supporting nonlinear, interactive storytelling and deep knowledge exploration through multidimensional narrative pathways.

This study explores the role of intertextuality in manuscript transmission using a Yule process model, extending previous birth-death approaches. Analysis of three major sets of medieval texts suggests that including speciation events better represents the heavy-tailed distribution of surviving witnesses per text.

This presentation introduces a scaffolded pedagogical guide to teach the practice of discovering, understanding, and evaluating DH tools and techniques that align with research objectives. It is founded on critical digital literacy skills for search processes and the context in which researchers develop new information.

This study examines ethical challenges in open science within Digital Humanities, focusing on balancing transparency, privacy, and compliance with data protection laws. By analyzing two COVID-19-related projects in Brazil and Portugal, it highlights informed consent practices, proposing adaptive governance models and strategies to harmonize openness with individual rights.

This paper presents a case study of web scraping for interdisciplinary research, highlighting both the practical implementation of these methods and strategies for overcoming similar challenges. We position web scraping as a crucial tool for ensuring data is accessible, not as a commodity, but in line with FAIR principles.

Some big emerging questions are about humans conversing ever more with AIs. For instance, how will we be affected? Yet, despite the volume of human/AI conversations, few public archives exist to preserve them. My paper introduces Preserving AI Voices, a public digital humanities project that creates such an archive.

Presents the pedagogical tool keylog.js, a minimal javascript-based tool that provides privacy-focused, client-side keylogging software served through a static website to address questions about the ethics, privacy, and accessibility of technologies and algorithms.

This contribution investigates the application of Handwritten Text Recognition (HTR) to the automated transcription of a multilingual historical manuscript.

This poster aims to present DoTS, a comprehensive and functional suite of tools for publishing corpora in compliance with the DTS specification, integrating backend, API responses, and frontend for the creation of adaptable websites.

This poster investigates how digital humanities (DH) scholars world-wide implement user experience (UX) practices in their projects. Through a survey, the authors will explore the integration of UX methodologies, identify barriers to adoption, and aim to promote more accessible, user-centered digital tools, ultimately broadening DH’s reach and engagement.

This research examines a corpus of contemporary migration novels to explore trauma. Using a mixed methodology, it investigates how these narratives depict human and non-human migrations, challenging anthropocentric environmental discourse while revealing the cultural imagination of climate change through recurring motifs, emotional arcs, and anticipatory memory.

The poster presents five strategies to produce more communal experiences in cultural-heritage software, exemplified using the Cultural Heritage Framework. The strategies are developed against the background of experts or 'rugged consumers' being the de-facto target audience of many editions, dictionaries, repositories and similar offerings.

We present Humanitext Antiqua, a system employing Retrieval-Augmented Generation (RAG) and advanced large language models (LLMs) to generate scholarly interpretations of classical texts. Using Plato’s Republic as a case study, the system integrates primary texts, commentaries, and secondary literature, addressing challenges in traditional referencing and text segmentation for enhanced academic research.

This paper introduces a machine learning approach to reassemble fragmented oracle bones, which are important materials for understanding early Chinese history. Specifically, we propose a model based on the Diffusion Model, a generative deep learning framework that has demonstrated remarkable performance in computer vision tasks in recent years.

This study proposes a comparison between three most popular chatbots: ChatGPT, Claude AI and Google Bard. Racial, ethnic and gender stereotypes were researched in the generated short stories. The stylo package for R and its function oppose() as well as the method of topic modelling were used in the study.

I have created a network that I have used to track influence, cashflow, and cruelty across 8 carceral institutions over approximately 50 years. In doing so, I can prove that there was an intricately weaved web of custodial cruelty among "fallen" girls and women in 19th Century Indiana.

Space, places, and spatial objects have been of interest for literary historical research for a long time. We take up this state of research by analyzing natural and artificial spaces, places, and spatial objects in nineteenth-century novels from Spain and Latin America.

This paper introduces a large-scale DH research project in Japan that the National Institute of Japanese Literature is undertaking over the next ten years. The aim of this project is to construct big data on Japanese pre-modern texts and to promote data-driven humanities.

This poster examines GitHub as a pedagogical platform in DH, analyzing its role in fostering coding literacy. By identifying pedagogical activities and practices, we explore how GitHub data can address gaps in training and inform the development of inclusive and effective programming pedagogies for the field.

The SSH Open Marketplace is a discovery portal which pools and contextualises resources for Social Sciences and Humanities research communities: tools, services, training materials, datasets, publications and workflows. This poster presents how this service can provide insights into the use of tools, methods and standards in the DH research communities.

The DALIA project aims to make open educational resources (OER) on data literacy accessible and interoperable. A knowledge graph is developed to link the materials, using the DALIA Interchange Format (DIF) to ensure transparency and interoperability. This poster focuses on picklists for DIF and invites feedback from the professional community.

This study analyzes OpenEdition’s server logs to uncover user navigation patterns across its platforms. Using methods like transition analysis, clustering, and topological modeling, it reveals platform fidelity, distinct user profiles, and shared interests. Future work aims to expand the scope with action-based analysis for deeper insights.

This poster presents the p2AFA corpus, a digital resource of Festival d’Avignon programs and playbills (1947–2024) for studying performing arts collaborations. Combining OCR, machine learning, and diplomatic transcription, it enables network visualization and historiographical analysis.

In this poster we will present the INTHEDATA project, in particular, the current state of knowledge and best practice in the area of cultural heritage and semantic knowledge model, the objectives of the project, its methodology by implementing the CIDOC-CRM standard and the first results.

We present our work on the development of machine learning classifiers trained to assess whether a given input image aligns with a specific beauty ideal. The work is part of our effort toward enabling large-scale computational research on beauty ideals, a subject that is both culturally significant and socially impactful.

This project aims to recover the cityscape of lost Muslim heritage in Belgrade by superimposing 200 destroyed mosques onto the modern cityscape at their original locations, using 3D-models in augmented reality, with textures based on the restored appearances of the mosques derived from old photographs, documents and sketches from travelogues.

This study aims to establish an effective information-provision method for humanities research, and as part of this effort, we developed an information retrieval system utilizing recently prominent technologies, LLM (Large Language Models) and RAG (Retrieval-Augmented Generation). This paper describes the developed system and its performance evaluation.

This poster highlights how Stadt.Geschichte.Basel created an Open Research Data Platform using CollectionBuilder. By applying minimal computing principles, the platform addresses challenges of accessibility, sustainability, and inclusivity in digital history. It provides adaptable, FAIR solutions that enhance interdisciplinary research, support marginalized perspectives, and foster long-term usability of historical data.

CLARIAH-ES is a Spanish national research infrastructure that strengthens research and facilitates innovative approaches within the digital humanities. By integrating language technologies, text analysis, cultural heritage, and multilingual resources, CLARIAH-ES offers a unique ecosystem for scholars interested in exploring the Spanish, Catalan, Galician, and Basque languages and cultures.

Google including Romani in their MT engine raises ethical concerns regarding linguistic preservation and cultural respect. Lack of transparency, poor translation quality, possible negative implications for language speakers are among the issues. It appears that the BigTech companies prioritize quality over quantity when it comes to support of lower-resource languages.

How can we sustainably build digital scholarship infrastructures that best serve their communities, encouraging co-ownership and input into their development? This poster examines the cooperative business model underpinning READ-COOP (https://readcoop.eu) and Transkribus (https://transkribus.org), an Automated Text Recognition platform, providing a blue-print for the establishment of responsible, democratic, cooperative digital infrastructures.

The ATRIUM project enhances access to digital research infrastructures in Arts, Humanities, and Social Sciences by improving services and creating a tailored curriculum for the research community. The poster showcases how, through a survey and workshops, ATRIUM integrates community feedback to bridge skills gaps and deliver impactful open training resources.

Scholarly Editing is an open-access, peer-reviewed journal that welcomes contributions that feature rare or marginal texts and small-scale editions for the discoverability of underrepresented stories and artifacts. This poster will introduce the journal’s purpose and present the journal’s strategies to ensure the longevity of its digital content.

O multilinguismo deveria afirmar-se nas Humanidades Digitais (HD). O objetivo deste trabalho é analisar se a produção científica em HD na Web of Science Core Collection o reflete. É um estudo bibliométrico com um corte transversal retrospetivo (2020-2024). Conclui-se que o inglês predomina, mas existe espaço para outros idiomas.

This project builds a web-based literary tour platform called ‘Memory of 518’, integrating literary works, factual records, and user-generated data related to the Gwangju May 18 Democratic Uprising. Using maps, 360-degree images, and user contributions, it documents and visualizes the fictional, historical, and everyday aspects of 518 Gwangju.

The proposed poster introduces a workflow for data extraction from historical maps into a structured format by manually digitising scanned maps with the OpenSource GIS software QGIS. We present our work in progress on a set of maps from our research in the UrbanMetaMapping project.

Digtal Humanities scholars need NLP tools to create new corpora of low-resourced languages, but such tools need to be trained on “non-existent” corpora creating a classic boot-strapping problem. We use the text from the Universal Declaration of Human Rights, along with a lexicon-based interative search strategy, to overcome this problem.

This study explores the visual representation of Africa in Chinese media (1950s-1980s), creating a digital archive and applying AI tools, including large language models and Contrastive Language-Image Pre-training (CLIP), for bidirectional text-image retrieval, offering fresh insights into Sino-African relations and cross-cultural visual studies.

NINDA (NINJAL Digital Archive) adapts Omeka S to manage linguistic resources, particularly for Japonic languages. It implements IIIF for multimedia content and OntoLex Lemon for lexical data structuring, supporting FAIR principles. The system handles annotated recordings, interlinear texts, and lexical databases, making linguistic resources more accessible to researchers and communities.

Textual scholarship aims to reconstruct and publish texts through critical apparatuses. The DSL-based Digital Scholarly Editions (DSE) method merges traditional editing with computational techniques, enhancing workflows and adhering to open science principles. The GreekSchools project exemplifies this approach, and the CoPhiEditor implements it as a software solution.

This poster will present a tool currently being developed for doing image vectorization of medieval script, called “Quil2Vec”. This tool to intended to expedite the production of image vectors as ground truth for multiple different text-based machine learning research applications.

The SCIROS project aims to enhance Open Science in the humanities and social sciences by tackling theoretical, practical, and infrastructural challenges with 6 international partners. By fostering interdisciplinary collaboration and sharing insights via the blog, the project supports the widespread adoption of OS practices.

This poster aims to present our ongoing work on developing the evaluation framework for open peer review assessment of non-traditional research outputs as a contribution toward maximising the quality and impact of Arts and Humanities research in Europe in the context of the Coalition for Advancing Research Assessment (CoARA).

After digital repatriation, cultural institutions often still remain digital replicas. This study compares metadata records of ancient Chinese paintings across various museums, revealing that interpretations in cross-cultural contexts are influenced by cultural backgrounds, and finally suggests improving original communities' control over digital replicas in legal, ethical, and technical aspects.

Die Sammlung und Bewahrung von marginalisierten Kulturgütern wie Privatbriefen unterliegen bisher keiner einheitlichen Archivierung und Standards. Das Dissertationsprojekt adressiert die Frage, wie die nachhaltige und dauerhafte Erschließung nach internationalen Standards im Spannungsfeld der FAIR-, CARE- und Open-Prinzipien gelingen und die maschinenlesbare und interoperable Digitalisierung des Kulturguts ermöglicht werden kann.

Our poster will present the partnership development grant “HumAInities: Exploring the Impact of AI on Humanities disciplines”, its goals and expected results. Our project seeks to understand the changes brought about by the impact of AI on the production and dissemination of knowledge within the humanities.

This study develops a Vedic Sanskrit OCR model to bridge the gap between text and image platforms.We fine-tunes TrOCR on Vedic, aligning images with texts using eScriptorium as a tool for creating groundtruth, suggesting a cyclic process to create text and image correspondences and to impove the performance of OCR.

This poster proposes a multimodal approach to historical research, utilizing HTR, NLP pipelines, and GIS in a user-friendly manner. By integrating advanced computational methods with traditional humanities research, we aim to create a model that can be replicated for other underrepresented regions and languages.

This poster explores the challenges and opportunities of digitizing late-antique literature, focusing on the Babylonian Talmud. By creating a comprehensive database of Talmudic stories, it aims to expand computational literary studies. The poster will discuss the methodological challenges involved in building this database, including text extraction, annotation, and modeling.

This study examines divergent language use in Russian state-controlled media and social media during the Russo-Ukrainian war using the WarMM-2022 corpus and Kullback-Leibler Divergence (KLD). KLD offers interpretability advantages over more opaque machine learning techniques allowing a deeper understanding of how propaganda techniques are linguistically construed and evolve over time.

A proposta visa desenvolver um modelo de gestão de acervos científicos alinhado com as diretrizes internacionais de Acesso Aberto (AA), especialmente considerando o Plano S e as transformações no sistema de comunicação científica.

This poster highlights three open source software packages I created in the programming language Rust. The packages include lemmatizing, readability, and stylometry algorithms, and were intentionally designed to create new resources to facilitate analysis of and engagement with multilingual and non-English languages in the Rust ecosystem.

Doing Literature is a web portal designed to collect and curate research contributions of the humanities in multimedia formats. It aims to develop an innovative framework that engages diverse audiences, thereby enhancing Digital Public Humanities and emphasising their collaborative character as well as their potential in knowledge creation.

This poster presents a user interface developed within the OntoVE project. The poster focuses on the search interface, built with the Sampo Model (Ikkala et al. 2022), and its search perspectives, which allow users to explore data through the faceted search paradigm (Tunkelang 2009).

This proposition showcases a Python library designed to interface with the text editor Obsidian to create a literary database of a corpus. We use Philip K. Dick's science-fiction as the exemplatory corpus, and showcase how classical Natural Language Processing can be used in computer-assisted literary hermeneutics.

We present an Open Information Extraction pipeline to identify and address knowledge gaps in Wikidata for underrepresented writers, using the Leksikon Fun Der Nayer Yidisher Literatur as a case study. Our approach benchmarks representation, assesses property alignment, and introduces resources to enhance digital humanities research on marginalized literatures.

This study presents a reproducible framework for publishing and reusing bibliographic metadata from GLAM, focusing on the European Literary Bibliography. It emphasizes Linked Open Data transformation, metadata enrichment, and computational reuse via Jupyter Notebooks. Key contributions include a framework, DH research scenarios, and tools enabling scholarly exploration of bibliographic collections.

Semantic modeling can play an important role in enhancing accessibility to the immaterial culture related to artifacts. To this end, we examine whether the domain standards CIDOC-CRM and LRMoo can express the interactions of an artwork with the contexts it traverses through a case study of XVII Century integrative restoration.

This study introduces a domain-specific ontology and digital archive for classical vocal music ephemera in South Korea. Addressing data fragmentation and inconsistent formats, it integrates Linked Open Data principles and visualization tools to ensure accessibility, cultural preservation, and analytical exploration across a decade of performance ephemera.

This paper presents a project undertaking the digitisation, enrichment, modeling and publication of modern and historical wine labels from the German Mosel region as witnesses of local cultural history using manual annotations and multimodal Large Language Models for enrichment and Linked Open Data for data modeling.

This panel of rhetoric, history, and Hispanic studies scholars aims to invigorate data advocacy research and education in the digital humanities by presenting and discussing the challenges and rewards of their work with three data-driven public humanities projects--a digital hate-tracking project, an online educational toolkit, and an online data repository.

This study utilizes phonotextual and cartographical perspectives to analyze Guangzhou Bamboo Branch Poetry, exploring emotional expressions and cultural landscapes. By examining textual features and Cantonese phonetics, we reveal the interplay of history, landscape, and local customs, highlighting the genre's significance in documenting urban life and cultural evolution.

This study explores scene change detection in 20th-century US-American romance fiction using manual annotations and automated methods. Manually annotated novels build the training data for fine-tuning an English BERT USE model, yielding promising preliminary results for automated text segmentation in computational literary studies.

This paper explores ways of representing the complex ways that landscapes can be described and how this changes over time drawing on the concepts of ‘picturesque’ and ‘wild’ in the English Lake District. It evaluates a range of approaches to landscape description and perceived nearness and how these changed over time.

Using Transformer-based models, this study contributes to an inclusive approach to understanding history, examining how males and females were depicted differently in medieval Chinese epitaph verses and which classical texts these verses resemble. The findings highlight a discourse shaped by patriarchal privileges, echoing the assertion that "the subaltern cannot speak."

This paper uses computational methods to analyse how Lady Augusta Gregory's translations of traditional Irish legends reimagined the original works for the purposes of aligning them with cultural nationalism and the Irish Literary Revivalist perspectives.

This paper examines references to Greco-Roman authors across disciplines in a corpus of 56,116 academic articles. Using digital methods, it identifies citation patterns, compares rankings with L’Année Philologique (bibliographic database of classical studies), and explores disciplinary differences.

Connecting Threads explores connections between South Indian weavers and Caribbean consumers by linking small and large textile collections and archives to enhance access to global fashion histories. Featuring a PostgreSQL database and interactive visualizations, the paper details its technical development, collaborative methodology, and impact on equity and accessibility in DH.

The paper presents a two-pronged approach for fostering access to and use of digital archival holdings. This approach combines public use of HTR technologies and public involvement in producing interpretative digital scholarship. The framework presented seeks to encourage civic engagement and dialogue around the holdings.

We introduce novel datasets and fine-tuned VLM and LLM models for OCR and word-sense disambiguation for Jawi (a writing system used historically for Malay). Our OCR system that outperforms previous solutions with a Character Error Rate (CER) of 8.66%, and a context-aware word sense disambiguation model that achieves 99.2% accuracy.

Most Nazi persecution memorials use physical maquettes for informing the historical site’s spatial organisation to visitors. In this paper we present the Future Memory Maquette Explorer from MEMORISE exhibition at Bergen-Belsen Memorial. It uses Augmented Reality technology to allow users to explore prisoner artworks, conveying the human dimension of history.

We aim to theorize the medieval manuscript page and its contents more holistically, using state of the art techniques to segment and describe the entire manuscript folio, for the purpose of creating richer training data for computer vision techniques, namely instance segmentation, and multimodal models for medieval-specific visual content.

Our project reimagines medieval sortes texts through virtual reality, combining textual scholarship with performative modeling. By situating these divinatory texts in immersive settings—monastic libraries, astrologers' laboratories, and taverns—we simulate their ritualistic and interactive nature. This approach bridges philology, media studies, and media archaeology, offering new insights into multimodal historical textuality.

Thanks to its microservices architecture, the eNDP application is designed as an access point to all the medieval resources available on Notre-Dame de Paris. We aim to present a method for valorizing and editorializing resources scattered across different research data repositories repositories, which takes advantage of their open APIs.

This paper reports on the outcomes of experimental work by the team involved with the Time Layered Cultural map (TLCMap) project funded by two Australian Research Council Linkage Infrastructure, Equipment and Facilities grants. It underlines the potential for digital mapping and digital humanities methodologies in historical and cultural heritage research.

The aim of the paper is to present the project “Ethnobotany of the Tambov region according to historical sources”. This is an ethnolinguistic and ethnobotanical study of the Tambov region traditional culture on the material of the lexical-semantic field “Plants”, with the use of the PhytoLex database.

This paper presents a framework for AI-driven heritage research at Silahtarağa Archive, a repository of historical documents and maps related to Istanbul's urban development. The authors outline their approach to digitization, data preparation, and AI methodology, enabling researchers to analyze and explore the evolution of Istanbul's urban landscape.

Kalpana explores the transformative role of digital technologies in reimagining museum spaces, focusing on immersive storytelling and experimental museology. Highlighting Global South perspectives, the project integrates AI-generated visuals and multimedia narratives to address inclusivity, accessibility, and decolonial aesthetics, offering innovative frameworks for cultural preservation and engagement in the digital age.

Accessibility and inclusivity are central to two University of Cologne projects compared here. Virtual Campus employs VR and AR for accessible campus navigation and cultural heritage engagement. ARTEST advances DH education and collaboration internationally. Both local and global initiatives for accessible, inclusive education are needed and benefit from mutual exchange.

This paper develops a critical ontological framework to map 19th-20th century performance histories in Asia's archipelagic regions. Introducing Archipelagic Performance Histories Ontology (APHon), the domain ontology captures fluid geo-social relations and touring practices, challenging nation-centric narratives through a structured yet flexible framework that represents complex cultural and historical data.

We investigated the use of virtual technologies to digitise and enhance cultural heritage (CH), aligning with Open Science and FAIR principles. Through case studies in museums, we developed reproducible workflows, 3D models, and tools fostering accessibility, inclusivity, and sustainability of CH. Applications include interdisciplinary research, educational innovation, and CH preservation.

This paper shows the necessity and motivation behind historicizing the power/knowledge embedded in LOD vocabulary systems. By utilizing CRMaaa, this paper presents a lightweight data model as a “quick fix” to augment existing vocabulary systems. This paper uses a particular case from a 19th c. painted panorama in Switzerland.

Sustaining open source software can be challenging. We discuss collaboration on the Universal Viewer (UV), software designed to display cultural heritage collections. We highlight methods including innovative, inclusive and multi-institution sprints. We showcase UV’s evolution, including accessibility and user experience enhancements, future plans and ways for others to contribute.

This study explores the emergence of digital history in Kazakhstan, Kyrgyzstan, and Uzbekistan, highlighting innovative projects, thematic focuses, and methodological shifts. Integrating literature reviews, interviews, and community events, it reveals infrastructural challenges, underscores postcolonial dimensions, and suggests that diverse, region-specific approaches can enrich global digital humanities discourse in meaningful ways.

This paper uses Chinese word vectors to develop a conceptual history of humanism and related keywords in a post-1945 modernist literary journal from Malaysia, and contributes to scholarship on digital multilingual practices, by asking how to accurately represent semantic and syntactic information from languages of non-Latin script in geometric spaces.

Using a historically informed approach, this contribution proposes a way to reconstruct historical ship 3D models, starting from their 2D drawings. It offers a study of the way ships were drawn, and uses this to charactyerise them as a parametrised problem.

This contribution proposes a multifaceted evaluation framework for assessing the performance of Large Language Models (LLMs) in encoding historical letters according to the TEI Guidelines, using the Joseph von Hammer-Purgstall correspondence edition as a case study.

We investigate the extent to which LLMs are able to learn a redefinition of a concept from literary studies, focalization, and apply it adequately to text examples. It shows that, with one exception, there are no statistically significant differences between the LLM output for prompts with and without the redefinition.

Aiming to study cross-cultural narrative patterns, this research develops a computational framework for extracting character features from English and Chinese fanfiction. By evaluating traditional Bert-based models and LLMs on tasks including character recognition, coreference resolution, dialogue and trait extraction, it provides insights into NLP tools' performance in characterization analysis.

We developed a system to automatically assign word sense tags to all content words in a substantial historical Japanese corpus, comprising over 20 million words. Our approach leverages a system based on Bidirectional Encoder Representations from Transformers (BERT), achieving an accuracy of 88.57%.

In this work, we propose a framework for extracting, annotating, attributing and visualizing dialogue characters in classic Chinese novels. We leverage interactive workflows to incorporate expert’s knowledge in the dialogue character extraction and attribution process.

This paper focuses on improving art-historical image search engines by combining Vision-language Models (VLMs) with knowledge graphs. The approach intends to enhance the interpretability and accuracy of search results by using triplets extracted from domain-specific knowledge with generative models. Providing this information to the user thus also increases the transparency of AI methods.

Cette communication présente une interface de recherche pour étudier la réception de l'art public au Québec avec les données générées par les utilisateur·rice·s de l'application MONA. L’outil développé optimise l'exploration et l’analyse des expériences artistiques recueillies pour contribuer de nouvelles perspectives sur l'interaction citoyenne avec l'art public.

In this study, we leverage two open datasets, respectively representing a dictionary of art symbols and scholars’ interpretations of ca. 400 artworks, to analyse how symbols and meanings vary in the art history hermeneutic discourse. Results show that the majority of scholar’s interpretations that could be aligned use conventional symbolism.

Our contribution examines methods for capturing spatial information from historical thematic maps depicting level of destruction during the second world war, focusing on the German city of Nuremberg (Nürnberg). We demonstrate three ways of capturing information from historical thematic maps, ranging from manual to semi-automated methods.

Using freely available data, this presentation tries to characterise the romance genre in the first half of the 20^th century and to map the part women screenwriters played in its evolution.

This study analyzed reception of ecological discourse in 20^th century US novel through computational criticism. We discovered a lexical family resemblance defined as “Pan-ecologicalness”, and implemented ecological discourse as a held-out example of genre changes throughout 20^th century.

This study presents an approach focused on natural habitats, plants and animals in German-language literature. To find out more about the aesthetic design, representation and distribution of ‘ecological codes’ we develop a classifier for animals, plants, and habitats in literary texts and apply it to 682 texts.

In a multilingual database of literary works, users will want to find a story's various versions. Therefore, we must conceptualize the threshold between narrative content (story) and its expression in language. While specially designed for evolving narrative traditions, our solution is grounded in the Functional Requirements for Bibliographic Records model.

To facilitate the description of versions and the creation of a version history, this contribution proposes a version assist system for digital editions. This system returns automatically generated change descriptions of changed resources that are comprehensible because each change is described as a purposeful, rule-based and contextualised action.

This proposal discusses the evaluation and editing of digital academic objects in digital humanities, emphasizing their epistemological value. It proposes a model based on Bhaskar’s publishing theory, focusing on academic rigor and technical precision. The goal is to enhance DH’s recognition and create a more accessible academic publications.

The contribution proposes a verb-class-based approach for the coarse-grained semantic classification of literary texts. Our annotations classify verbal phrases based on the semantic class of their main verb. Despite potential quality issues at the micro level, we demonstrate that this approach can yield valuable insights at the story level.

This study uses statistical word frequency analysis to explore the Groot Verseboek anthology of Afrikaans poetry, examining dominant themes, stylistic trends, and shifts in socio-historical context. By combining Digital Humanities and literary analysis, it reveals how canon formation reflects cultural and ideological values, offering new perspectives on Afrikaans literature.

This presentation showcases a case study from VERITRACE, an ERC project using digital tools to identify the influence of ancient wisdom traditions on early modern science, highlighting connections between ancient texts and the works of Kepler, Galileo, and Descartes through text matching and semantic analysis, revealing multilingual traces of influence.

The project From Parchment to Computer offers courses on Textual Criticism and digital scholarly editing, blending theory and practice. It aims to train participants in creating digital editions, promote Digital Philology, enhance critical understanding of digital technologies, and contribute to democratizing access to these tools while advancing DH in Portugal.

This contribution discusses an innovatively stratified collection of German language data, ranging from informal spoken interactions to formal written texts. It highlights methods for analyzing linguistic patterns using natural language processing, with a particular focus on discourse markers and a machine learning model for identifying them across diverse communicative contexts.

Research Infrastructures (RIs) in the humanities actively shape our understanding of the world. Using examples of bibliographies and corpora, this paper examines how methodological choices in building RIs in digital literary studies influence representation, advocating for open infrastructures to ensure inclusivity and a more nuanced understanding of the literary landscape.

With the rise of LLM and the increasing computational expense of AI, humanists will increasingly turn to high-performance computing (HPC). This can be an alienating pivot for many researchers. As a Research Data Facilitator with a decade of HPC experience, I will present models for effectively integrating humanists into HPC.

The paper examines key features of research infrastructures in humanities and cultural heritage that support open science, focusing on federated RIs as marketplaces connecting diverse networks. It discusses the technological foundation and API orchestration for DH workflows within the H2IOSC MarketPlace, highlighting contributions of the OPERAS node to this project.

As the triad of publishing a paper, data and code poses challenges for the comprehensibility, reproducibility, and accessibility of the research, we present our approach towards a "modularised open infrastructure for research publications” in which a publication is accompanied by modules facilitating e.g.reproduction or result investigation.

We investigated the semantic and rhetoric imagery of Kafka’s novel The Trial through three illustrated editions of the text. Using image analysis techniques and examining the relationship between images and corresponding texts, we found these illustrations more closely associated with sentences than chapters and uncovered their artistic and hermeneutic nuances.

This ongoing project provides a novel workflow for studying Weimar Germany’s political culture. By integrating text-hermeneutic investigation with quantitative digital analysis techniques, it enables new insights into historical newspaper discourses on democracy. The project, therefore, enhances historical newspaper research and contributes to the understanding of interwar Germany.

We hypothesize narrative volatility (shifts in sentiment between chunks of text) has an effect on appreciation and thus on ratings of fiction. We describe how we compute volatility and show its distribution over genre. We explain how we will use the result to test the hypothesis.

We analyze how often Indian English and German novels (2000–2024) refer to information technologies (IT), reflecting demographic, cultural, and societal differences. We use a word-list approach and and a large language model. The llm-based approach works well, but the result doesn't confirm our hypothesis that there is a significant difference.

This paper showcases layout analysis and OCR to make an under-researched, 120 year old crime journal accessible. It then uses a variety of text analysis tools for distant reading, exploring how crime was addressed at the time.

Our submission reviews 20 digital interfaces for CH data from 2015 to 2024. It finds 6 common objectives stated by the authors of the reviewed use cases but highlights that not all stated objectives are equally well evaluated.

This short presentation examines the overlaps between technical and professional communication (TPC) and digital humanities (DH) by using TPC content analysis methods to examine the interfaces of 100 DH project websites. It describes common DH web design features and offers insight regarding the development of accessible and sustainable DH projects.

The CLARIAH-EUS-gArA project aims to enhance DH research by developing a trustworthy conversational assistant for Basque news using RAG and Latxa, a Basque LLM. It integrates AI and LT to address misinformation, verification, and reliability, thereby providing accurate, up-to-date responses in Basque to aid researchers in fact-checking and accessing sources.

This study evaluates Graph RAG’s applicability to the humanities, focusing on Caesar’s Gallic Wars, volume 1. A knowledge graph was constructed using LLMs, enabling the retrieval of semantically structured data. The results highlight the potential of enhanced knowledge graphs for broader applications, emphasizing evaluation methods and expert-driven graph development.

Non-coding tools have expanded accessibility in digital humanities, empowering researchers without programming skills to perform data-driven analyses. However, there are currently few tools to assist non-coders with converting and cleaning data. This paper presents a Shiny application for data preprocessing positing that similar small-scale solutions could help bridge this gap.