TermTrends 2025 is a half-day workshop co-located with LDK 2025 that aims to provide a discussion forum on the theoretical and methodological approaches for the integration of language resources, and specifically terminological resources, published following Linked Data principles, into the optimisation of LLM pipelines. In particular, this event focuses on how terminological resources can be leveraged to enhance the performance of LLMs at different stages of their lifecycle, from pre-training and fine-tuning to evaluation and application. By identifying points of convergence between Linked Data-based terminological data and LLM processes, we seek to explore strategies to make these resources more interoperable, reusable, and impactful.

To this end, the workshop intends to explore the advantages and challenges of leveraging Linked Data-compliant language resources in the context of LLMs. This includes examining how existing standards for representing terminological data, such as TermBase eXchange (TBX) format, and Semantic Web models like SKOS and Ontolex-lemon, can be adapted or extended to meet the needs of LLM development. In addition, the workshop aims to discuss innovative ways to integrate language resources published as Linked Data into LLM workflows to address issues such as domain adaptation, bias mitigation, and the interpretability of model outputs.

Furthermore, the workshop will examine the potential of LLMs themselves as tools for generating or enriching terminological resources published as Linked Data. This includes exploring how LLMs can support the automatic extraction, structuring, and linking of terminological data to Semantic Web standards, enhancing their quality and accessibility for downstream applications. By bridging the gap between LLM capabilities and Linked Data methodologies, the workshop aims to identify synergies that benefit both fields.

REGISTRATION: The registration will be handled by the main LDK 2025 registration web page. All the details will be published there.

DATE and VENUE: The workshop will take place on 9th September 2025 and it will be held as a hybrid event. The venue for the Workshops and Tutorials will be Università degli Studi di Napoli L'Orientale - Palazzo del Mediterraneo.

Patricia Martín-Chozas, Universidad Politécnica de Madrid (Spain): She works as an Assistant Professor at Universidad Politécnica de Madrid (UPM) and is member of the Ontology Engineering Group at the same University. Her research has been oriented to the generation and representation of terminological resources by means of Semantic Web technologies. She is currently involved in several research projects (INESData, TeresIA, Innovatrad), where she combines those research areas. Her next research steps are focused on the exploitation of terminological resources published as Linked Data to improve the performance of Large Language Models.

Elena Montiel-Ponsoda, Universidad Politécnica de Madrid (Spain): She is an Associate Professor of Applied Linguistics at Universidad Politécnica de Madrid (UPM), Spain, and a member of the Ontology Engineering Group at the same University. Her main research interests are in the common ground between Terminology and Ontology Engineering. Her research has focused on the development of models to enrich ontologies with multilingual information and to expose terminologies and other language resources as linked data. She is currently involved in several research projects (INESData, TeresIA) that explore the automatic generation and conversion of terminological data, as well as their influence in the training or fine-tuning task of Large Language Models.

Sara Carvalho, University of Aveiro (Portugal): She is a member of the Languages, Literatures and Cultures Research Centre of the University of Aveiro (CLLC-UA), as well as a collaborator at the Linguistics Research Centre of the Universidade NOVA de Lisboa (NOVA CLUNL). In addition, she is a member of the ISO/TC 37 and of the Portuguese mirror committee "CT 221 – Terminologia, Língua e Linguagens" at the Portuguese Institute for Quality. Her research interests include terminology, lexicography, and knowledge organisation (especially as regards the connection with ontologies), plus the modelling and sharing of terminological resources as linked data, particularly in the biomedical domain.

Federica Vezzani, University of Padova (Italy): She holds a PhD in terminology and is an assistant professor at the Department of Linguistic and Literary Studies of the University of Padova, Italy. She is a member of the ISO/TC 37 "Language and Terminology" and of the Portuguese mirror committee "CT 221 – Terminologia, Língua e Linguagens" at the Portuguese Institute for Quality. Her main research interests are terminology, specialised translation, and technical communication. In particular, she focuses on the management of multilingual terminology according to ISO standards, and she has developed the FAIR terminology paradigm for the optimal organisation of findable, accessible, interoperable, and reusable terminological data.

TO BE ANNOUNCED

For more information about the tutorial, please contact us by email: patricia.martin@upm.es, elena.montiel@upm.es, sara.carvalho@ua.pt, federica.vezzani@unipd.it