TRIFECTA: Capturing Identity, Change, and the Long Tail in Knowledge Graphs
Description: The TRIFECTA project is an ERC Consolidator project on the intersection of Semantic Web, Natural Language Processing, and digital humanities. The project aims to address three core knowledge extraction and modelling challenges: identity, change, and the long tail. Two use cases are investigated in the project: maritime history and food history. Humanities use cases are highly suited to taking on these challenges as often data is available from different sources spanning a longer period of time. Such collections of resources provide examples of the three core problems TRIFECTA aims to address.
Presenters: Marieke van Erp, Gauri Bhagwat, Izzie He, Teresa Paccosi, Katrina Slebos Perez, Veruska Zamborlini and Jiaqi Zhu
Website: https://trifecta.dhlab.nl
REEVALUATE: Framework for safe, open, collaborative and inclusive digitisation and management of cultural heritage
Description: REEVALUATE is a Horizon Europe project in the Cultural Heritage (CH) domain that provides an actionable framework of guidelines, recommendations and technical enablers for the digitisation, management and reuse of cultural artefacts. At the core of the framework sit the ARTKB knowledge graph (currently∼8 M RDF triples and∼147,500 linked digital artefacts from four CH institutes and Wikidata), structured according to the Cultural Artefact’s Contextual Ontology (CACAO), and the REEVALUATE marketplace built on top of it. Together they form the integration point for seven AI-based enablers covering prioritisation, contextualisation (image, text and audio), human-driven contextualisation, collaboration, creative reuse, context validation, and IPR management. All enablers are released as open-source packages and are being exercised in three real-world pilots covering the Fashion, Advertising and Tourism Cultural and Creative Industries. As of October 2025 all enablers have reached at least TRL 5, with most already pilot-tested (TRL 6).
Presenters: Ruben Peeters, Xuemin Duan, Giacomo Blanco, Tommaso Monopoli, Federico D'Asaro, Panagiotis Stalidis, Giuseppe Rizzo and Anastasia Dimou
Website: https://reevaluate.eu/
The CRC 1625 Semantic Layer: Establishing a Knowledge Graph for the Design of Complex Solid Solution Surfaces
Description: The CRC 1625 project A06 addresses the challenge of establishing a Knowledge Graph (KG) that provides a unified semantic representation of the data used and generated within a Collaborative Research Centre (CRC) dedicated to the design of Compositionally Complex Solid Solution Surfaces (CCSS). The KG integrates heterogeneous data originating from high-throughput experiments, simulations, theoretical and statistical models, machine learning predictions, and domain expertise, to enable the systematic discovery of novel catalytic materials. The project investigates novel techniques for representing CCSS in a compact yet meaningful way, constructing the KG from heterogeneous sources, incorporating complex expert knowledge through rules, and efficiently querying the resulting large-scale, heterogeneous KG. By combining ontology engineering, knowledge graph construction, human-in-the-loop approaches, and query optimization, A06 aims to deliver both a semantic database describing CCSS and a semantic layer integrating the diverse data sources of the CRC. The outcomes of the project will contribute actionable knowledge to support experts and data-guided approaches in the discovery of new materials with desired catalytic properties.
Presenters: Maribel Acosta and Samuel García Vázquez
Website: https://www.ruhr-uni-bochum.de/crc1625/projects/projecta06.html.en
SOEL: Supporting Ontology Engineering with Large Language Models
Description: The influence of Large Language Models (LLMs) is currently boosting and reshaping current practices in numerous fields, for example computer science. Due to their high capabilities in natural language understanding, their impact has been particularly evident in the field of Natural Language Processing. However, other areas of Artificial Intelligence, like knowledge representation and ontology engineering, may also benefit from generative AI technologies. Developing ontologies is still a challenging and time-consuming task even though many tools have been developed by the scientific community to assist developers, for example, for formalizing tests to assess requirements, creating human-readable documentation for ontologies, ontology assessment, etc. In general, a significant manual effort is still required from ontology developers to conceptualize, reuse, and validate existing ontologies. To alleviate such ontology engineering tasks, several works have arisen proposing the use of LLMs to support the generation of competency questions, the conceptualization of models from text, the identification of concept alignments, etc.). However, the tasks addressed in previous works are usually defined in an heterogeneous manner, with different scope, inputs, expected outputs and metrics. There are no reference evaluation tasks, datasets or benchmarks that help compare existing work against future efforts. To address this situation, SOEL aims to explore the application and adaptation of (Large) Language Models in Ontology Engineering tasks by creating open reference datasets, defining evaluation tasks and comparison benchmarks through adapting state-of-the-art LLM models. Although SOEL’s approach is independent of a specific domain, it is planned to assess the results in three specific domains in IoT (smart buildings), Linguistics (definition of terminologies), and Open Science (description of scientific artifacts and experiments).
Presenters: María Poveda-Villalón and Daniel Garijo
Website: https://w3id.org/soel
EAD: Environmental Anomaly Detection and Health Impacts for PFAS Risk Analysis
Description: Environmental pollutants such as Per- and Polyfluoroalkyl Substances (PFAS) pose major challenges for both ecosystems and public health due to their persistence and toxicity. The EAD project develop approaches and methods that facilitate the analysis of data on pollutants impacting the environment and water resources. Our focus is particularly on analyzing data at national and European levels regarding PFAS. The project uses the PFAS Data Hub (PDH), a research project run by the CNRS Humanities & Social Sciences, to detect anomalies, identify contamination patterns, and assess potential health impacts. By combining ontology-based modeling, hybrid anomaly detection approaches, and visual analytic, EAD contributes to the development of explainable and interoperable solutions aligned with the One Health paradigm.
Presenters: Lylia Abrouk, Davide Di Pierro and Danaï Symeonidou