This is a companion proceeding volume to the 21st International Conference on Semantic Systems, SEMANTiCS 2025, which took place from September 03-05, 2025, in Vienna. It collates all accepted posters, demos, and tutorials, as well as the proceedings from the eight workshops co-located with SEMANTiCS 2025. SEMANTiCS is the annual meeting place for professionals who make semantic computing work, understand its benefits, and encounter its limitations. Every year, SEMANTiCS attracts information managers, IT architects, software engineers, and researchers from organizations ranging from research facilities and NPOs through public administrations to the largest and/or most innovative companies in the world. Conference participants learn from top researchers and industry experts about emerging trends and topics in the wide area of semantic computing. The SEMANTiCS community is highly diverse; attendees have responsibilities in interlinking areas such as Artificial Intelligence, knowledge discovery and management, bigdata analytics, e-commerce, enterprise search, technical documentation, document management, business intelligence, and enterprise vocabulary management. In Section 2 we discuss the accepted posters and demos, whereas in Section 3 we briefly introduce the co-located tutorials and workshops. Finally, in Section 4 we present the program committee of the posters and demos track.
The Posters and Demos track serves as a platform for presenting ongoing research, innovative applications, prototypes, and early-stage results. This track enables participants to share novel concepts and obtain feedback in an interactive environment. In 2025, the SEMANTiCS conference accepted 11 posters and 9 demos. The accepted submissions represent a diverse range of current research in knowledge graphs, semantic technologies, and foundational models. The program includes papers on the automatic construction and generation of knowledge graphs from domain-specific datasets, including Internet of Things (IoT) platforms, renewable energy systems, and configuration data. These works emphasize the integration of scalable and interoperable knowledge. Additional themes include semantic enrichment and data governance, with research focusing on change management in vocabularies, the FAIR data principles, and digital product passports. Several papers examine the intersection of knowledge (S. Vahdati)
CEUR
ceur-ws.org graphs, large language models, and artificial intelligence methods, focusing on hybrid approaches, retrieval-augmented generation, and cognitive complexity frameworks to enhance reasoning and information retrieval. Usability and exploration are also addressed through semantic search engines, ontology visualization tools, and user interfaces for SPARQL endpoint and graph query interaction. The collection also features research on infrastructure and performance optimization, including edge–cloud orchestration, declarative API generation, query rewriting for security, and evaluations of graph engines. Domain-oriented applications include energy co-simulations, archival data curation, and agroforestry research. Collectively, these contributions demonstrate a balance between foundational research, applied innovation, and system-level engineering within the semantic technologies community.
The workshops and tutorials track allow any organization or project to promote any of the 2025 SEMANTiCS research topics and gain increased visibility. 2025 workshops and tutorials are incubators for industrial and scientific communities that form and share a particular research and development agenda, and they provide a forum for presenting contributions and findings to a diverse and knowledgeable community. This year SEMANTiCS accepted eight workshops and two tutorials as part of the conference program. We briefly describe them below: • The Semantics for Transport (Sem4Tra25) workshop aims to advance the Mobility and Logistics Data Space using Semantic Web, Linked Data, and Knowledge Graph techniques. Use cases include trip planning, booking logistics, and supply chain insights. Ideal for researchers and practitioners transforming passenger and freight transport operations. • The Second Knowledge Graphs and Neurosymbolic AI (KG‑NeSy) Workshop explores the synergy between symbolic Knowledge Graph approaches and Neurosymbolic AI/ML. Contributions on graph construction, learning, interpretability, and hybrid AI systems are invited. • The Developers Workshop (SemDev) is designed to bridge academia and industry in semantic software development. Participants will present tools, showcase live demos, and discuss open implementation and research problems. • The Symbolic & Generative AI for Science (SymGenAI4Sci) Workshop investigates how generative and symbolic AI models can support scientific discovery from hypothesis formation to data interpretation, while addressing interpretability, reproducibility, and ethical concerns. • NeXt‑Generation Data Governance Workshop (NXDG) focuses on privacy-aware data spaces, compliance with EU regulations (GDPR, Data Act, AI Act, etc.), and semantic policy frameworks like ODRL and DPV. Key for technical, legal, and societal stakeholders in the data economy. • The Semantics for Transparency in Industrial Systems (SENTIS) Workshop targets the growing complexity of industrial systems at the IT–OT–AI intersection. Showcases the role of semantic models in enhancing explainability and transparency in distributed, software‑driven ecosystems. • The Workshop on Users and Knowledge Graphs (UKG) centers on the human side: how users interact with, perceive, and understand knowledge graphs. Topics include usability, query interfaces, user evaluation methods, and enhancing KG accessibility. • The Scaling Knowledge Graphs in Industry: LLMs meet KGs (SKGI) Workshop Examines industrial applications at the intersection of Large Language Models (LLMs) and Knowledge Graphs. Themes include Graph‑based RAG, scalable KG creation, energy‑eficient AI, and explainable, user‑centered KG systems.
As for tutorials, SEMANTiCS 2025 included the following: • Making Knowledge Graphs AI‑Ready: Quality Assessment & Improvement in GraphRAG is a hands-on tutorial focused on evaluating and enhancing knowledge graph quality for reliable GraphRAG pipelines. The tutorial will cover failure patterns, modeling pitfalls, and criteria to ensure semantic reliability and retrieval efectiveness. • Streaming & Cross‑Environment Pipelines with RDF‑Connect is an interactive tutorial on using RDF‑Connect to build multilingual, provenance-aware, streaming data pipelines. Participants will build an example pipeline (e.g. a knowledge graph from weather forecasts) across Python, JavaScript, and Java, with transparent provenance tracking via the W3C Provenance Ontology (PROV‑O).