Invited Talks

With the rapid advancement of large language models, generative AI has penetrated academic research extensively, bringing efficiency gains as well as severe challenges to academic authenticity. Reliable detection of AI-generated academic text has become an urgent research issue. Existing methods mainly target general texts and perform poorly on professional standardized academic writing, and their robustness declines rapidly with evolving LLMs. This study proposes an AI academic text detection method from the fundamental perspective of text complexity. We construct a multi-dimensional complexity indicator system and model joint feature distributions via Gaussian Mixture Model. Evaluated on datasets covering GPT-3.5-turbo, GPT-4o and GPT-5, our method achieves leading performance compared with 12 baselines, shows strong cross-model adaptability and remains stable under adversarial paraphrasing attacks.

Cataloging rules as content standards and encoding formats based on the LRM conceptual model should be closedly interrelated to ensure effective representation of resources. Numerous studies have examined the mapping among these standards, including mappings from conceptual models to cataloguing rules, from cataloging rules to encoding format, and from conceptual models to encoding formats. Examples include studies on mappings between LRM and RDA, RDA and BIBFRAME, as well as between LRM and BIBFRAME. Last year, the Korean Cataloging Committee developed KCR5 (previous KCR4) based on the original RDA. Accordingly, this study aims to evaluate the suitability of KCR5 for the exchange and sharing of bibliographic data within the BIBFRAME and RDA frameworks by mapping KCR 5 to both RDA and BIBFRAME.

This talk explores how people experience, interpret, and construct meanings of cultural heritage within a heritage metaverse environment. Drawing on a mixed-methods study, it examines how virtual heritage settings, interactive activities, and social engagement contribute to participatory heritage interpretation in digital environments. By analysing the relationships among these elements, the talk highlights design insights for developing user-centred heritage metaverse experiences. It further discusses how participatory interpretation can contribute to more inclusive approaches to heritage communication, knowledge organisation, and the representation of cultural meanings in digital environments.

This talk is a follow-up to a panel from the last DCMI to give updates on the recent activities. In March 2026 the Data Documentation Initiative (DDI) technical document DDI Common Core was successfully published as an ISO standard, with significant support of members of the Dublin Core community. The presenters will briefly introduce the DDI product suite (including the Common Core), the embedding of DC in DDI, the process of the ISO/PAS certification of DDI and DC’s role in the overall process.

Linked Open Data (LOD) can be published on the Semantic Web as semantic mark-up embedded in HTML documents or as Knowledge Graphs (KG) in SPARQL endpoints, on top of which applications are developed. KGs typically share the same entities, such as persons, organizations, and places, which means that the data in different KGs about the entities can complement (enrich) or contradict each other, if only we knew what entities are the same in different KGs. For this purpose, entity alignment systems are needed. Examples of such systems include, e.g., the Virtual Authority File VIAF.org of national libraries around the world, the international Linked Open Data Cloud, Linked Open Vocabularies (LOV), and Europeana. This talk presents a new data alignment service and semantic portal "SampoSampo - Connecting Everything to Everything Else" implemented on top of a cloud of 26 related Cultural Heritage (CH) KGs and systems in different CH domains, including biographies, genealogy, fiction literature, historical letters, art, military history, parliamentary discussions, and historical opera performances. The novelty of SampoSampo is to show in practise how a data alignment system can be used for five practical use-cases in Digital Humanities research (DH): 1) Using a LOD alignment service to enrich data in other Cultural Heritage LOD services and applications. 2) Providing global search, browsing, and data-analytic tools over a cloud of KGs. 3) Searching and checking data quality issues based on a cloud of KGs. 4) AI-based knowledge discovery in a LOD cloud based on relational search and Large Language Models. 5) Enhancing data literacy by data-analyses over a KG cloud. SampoSampo was published openly on the Semantic Web in January 2026.

This talk will highlight key tools and features of International Image Interoperability Framework (IIIF) that enable, empower, and advance the content of digital libraries. The increasing importance of AI in the cultural heritage sector is also refactoring the use cases for IIIF. Examples of ongoing collaboration of the IIIF and AI4LAM (Artificial Intelligence for Libraries, Archives, and Museums) community will be featured. Also covered will be the mature standard IIIF functions such as deep zoom, annotations, and multiple asset reference (e.g. image comparisons). These elements of the IIIF standard will be highlighted to show how new tools and platforms are being developed using IIIF. Chief among these will be Allmaps [1] (which allows users to geo-reference a IIIF-enabled historic map on a modern map facilitating overlaying of maps of different periods for comparison and also allowing access to maps of locations from institutions across the world. An update on the status of the IIIF Presentation API 4.0 which further facilitates use of 3D content will also be covered [2]. The growing IIIF community will also be featured. The IIIF Content Commons, a developing concept that will enhance the findability of IIIF resources will also be explored. IIIF [3] is a set of open standards for delivering high-quality, attributed digital objects online at scale. It’s also an international community developing and implementing the IIIF APIs. IIIF is backed by a consortium of leading cultural institutions and a global and engaged community [4]. IIIF is a way to standardize the delivery of images and audio/visual files from servers to different environments on the Web where they can then be viewed and interacted with in many ways IIIF makes these objects work in a consistent way. That enables portability across viewers, the ability to connect and unite materials across institutional boundaries, and more. This portability is one of the reasons IIIF is an important standard in digital humanities research allowing open access collections to be used as research data and for research software to be written once and used with many collections. [1] Allmaps a map geo-referencing tool https://allmaps.org/ [2] IIIF Specification roadmap. https://iiif.io/news/2025/08/11/roadmap/ [3] International Image Interoperability Framework IIIF https://iiif.io [4] IIIF Specification roadmap. https://iiif.io/news/2025/08/11/roadmap/

Metadata has long been recognized as a fundamental component of research data management and reuse. However, the extent to which metadata practices facilitate data reuse remains insufficiently explored. This study investigates metadata practices in research data management across scientific disciplines. Data were collected from the Data Citation Index.. Five categories of metadata elements were identified and analyzed. The findings reveal that descriptive metadata and author-related information were the most prevalent across datasets, whereas persistent identifiers such as DOIs and ORCID iDs were used infrequently. These results contribute to a deeper understanding of metadata practices and highlight key metadata elements that may support the reuse of research data.

Ensuring long term access to digital materials is no longer optional—it is a core responsibility for every institution stewarding digital content. Karin Bredenberg, chair of the PREMIS Editorial Committee, offers a focused introduction to digital preservation, highlighting the critical role of preservation metadata and the PREMIS standard in sustaining the authenticity, reliability, and usability of digital objects over time. It demonstrates why preservation cannot rely on a single metadata standard and Bredenberg will argue for a strategic, coordinated use of multiple standards to meet the complex demands of long term stewardship. Bredenberg will clarify how preservation metadata complements descriptive metadata, dispelling the common misconception that using several standards leads to unnecessary duplication. Instead, it will be shown how intentional overlaps strengthen interoperability and support preservation workflows. Participants will gain a clearer understanding of why preservation metadata matters, how PREMIS fits into a broader metadata ecosystem, and how adopting these practices can significantly enhance the long term value of their digital collections and their continued trustworthiness and reusability.

This presentation shows how the domain-specific “Sampo framework”, here ParliamentSampo for parliamentary speeches and prosopography [1], based on the domain-agnostic Sampo model [2] and Sampo-UI interface model [3-4], can be adapted to new application cases. As an example, our focus here is to show how to publish as a FAIR Linked Open Data (LOD) service the assembly minutes of the League of Nations [5], the predecessor of the United Nations (UN), and to publish next the General Assembly minutes of the UN. As an application on top of the LOD service, the portal League of Nations (1920-1946) Sampo [5] was published on the Web in March 2026. It can be used to answer questions, such as: What have the representatives of Japan talked about? How did the territorial focus of the League of Nations (LoN) shift in time? In what roles and capacity did Fridtjof Nansen, the Polar explorer, participate in LoN meetings? More information about the underlying Minutes of Multilateralism initiative: https://seco.cs.aalto.fi/projects/minutes/ [1] Eero Hyvönen, Laura Sinikallio, Petri Leskinen, Senka Drobac, Rafael Leal, Matti La Mela, Jouni Tuominen, Henna Poikkimäki and Heikki Rantala: Publishing and Using Parliamentary Linked Data on the Semantic Web: ParliamentSampo System for Parliament of Finland. Semantic Web, vol. 16, no. 1, 2025. DOI: 10.3233/SW-243683. https://journals.sagepub.com/doi/full/10.3233/SW-243683 [2] Eero Hyvönen: Digital Humanities on the Semantic Web: Sampo Model and Portal Series. Semantic Web – Interoperability, Usability, Applicability, vol. 14, no. 4, pp. 729-744, IOS Press, 2023. https://journals.sagepub.com/doi/full/10.3233/SW-223034 [3] Heikki Rantala, Annastiina Ahola, Esko Ikkala and Eero Hyvönen: How to create easily a data analytic semantic portal on top of a SPARQL endpoint: introducing the configurable Sampo-UI framework. VOILA! 2023 Visualization and Interaction for Ontologies, Linked Data and Knowledge Graphs 2023, pp. 28-39, CEUR Workshop Proceedings, Vol. 3508, October, 2023. https://ceur-ws.org/Vol-3508/paper3.pdf [4] Esko Ikkala, Eero Hyvönen, Heikki Rantala and Mikko Koho: Sampo-UI: A Full Stack JavaScript Framework for Developing Semantic Portal User Interfaces. Semantic Web – Interoperability, Usability, Applicability, vol. 13, no. 1, pp. 69-84, January, 2022. https://journals.sagepub.com/doi/full/10.3233/SW-210428 [5] Petri Leskinen, Eero Hyvönen, Alexandre Lionnet, Blandine Blukacz-Louisfert, Pierre-Etienne Bourneuf, Davide Rodogno, Grégoire Mallard, and Florian Cafiero: A Linked Open Data Service and Semantic Portal to Study the Assembly Minutes and Prosopography of the League of Nations (1920–1946). The Semantic Web: ESWC 2026, Dubrovnik, Croatia, May 10 - 14, 2026, Proceedings, Springer-Verlag, 2026. In press. https://seco.cs.aalto.fi/publications/2026/leskinen-et-al-lon-2026.pdf