This second edition expands the scope to explicitly include Large Language Models (LLMs) and their intersection with Knowledge Graphs. This reflects the increasing demand in industry for hybrid symbolic-neural systems that combine the robustness and structure of KGs with the flexibility and generative power of foundation models.

We are particularly interested in research and industrial innovations that explore how symbolic and neural approaches can be combined to create scalable, explainable, and trustworthy AI systems suitable for real-world applications. 

To foster deep and practical discussions around the integration of Knowledge Graphs (KGs) and Large Language Models (LLMs), we invite contributions that address both foundational and applied challenges.

🗓️ Half-day: September 03 - 05, 2025 | 📍 Vienna, Austria | 📚 Co-located with SEMANTiCS 2025

🔗 Submission portal: Open Review

🔗 Previous edition: SKGI 2024

Author Guidelines

📄 Submission Guidelines

• Long papers (8–12 pages): Research, In-use systems, or comprehensive LLM-based systems evaluations

• Short/position papers (4–6 pages): works in progress, early result exposure, position statements, emerging systems.

• All submissions will undergo peer-review by at least two research-community active members.

• Accepted papers will be submitted to  CEUR-WS proceedings. 

• All deadlines are set for 11:59 pm, Anywhere on Earth time (UTC-12)

• Presentation of accepted work at the workshop is mandatory.

• Page limits exclude references.

• Papers must be submitted in PDF format using the CEUR-WS single-column template. Templates are available  from CEUR-WS template .  

• Submission Platform: Open Review SKGI 2025

TOPIC OF INTEREST

We invite contributions from both academia and industry across a wide range of topics including:

• GraphRAGs Oriented Topics

• Graph-based Retrieval-Augmented Generation (GraphRAG)

• Using LLMs for semantic data ingestion and KG enrichment

• Evaluation frameworks for hybrid LLM-KG systems

• LLM hallucination mitigation using structured knowledge

• Real-world industrial case studies of KG/LLM integration

• Scalable reasoning and inference over symbolic-neural pipelines

Other LLM+KGs applications in the Real-World Settings

• Scalable construction and maintenance of KGs

• Streaming and temporal data in knowledge graphs

• Energy-efficient and sustainable KG-based AI systems

• Trustworthy AI: explainability and traceability through KGs

• User interfaces and human-in-the-loop systems for KGs