DataScientia Member Davide Cavicchini Receives Best Student Paper Award at IJCAI-ECAI 2026
DataScientia is proud to celebrate the success of community member Davide Cavicchini, whose research paper “KoRe: Compact Knowledge Representations for Large Language Models” received the Best Student Paper Award in the Research Track of the joint GenAIK–NORA 2026 Workshop, held alongside IJCAI-ECAI 2026 in Bremen, Germany.
The award recognizes the work of Davide Cavicchini, Fausto Giunchiglia, and Jacopo Staiano for their contribution to the emerging intersection of Generative AI and Knowledge Graphs. The paper was selected for its quality, originality, technical contribution, and presentation during the workshop.
Here is the paper’s abstract:
Modern Large Language Models (LLMs) have shown impressive performances in user-facing tasks such as question answering, as well as consistent improvements in reasoning capabilities. Still, the way these models encode knowledge seems inherently flawed: by design, LLMs encode world-knowledge within their parameters. This way of representing knowledge is inherently opaque, difficult to debug and update, and prone to hallucinations. On the other hand, Knowledge Graphs can provide human-readable and easily editable world knowledge representations, and their application in knowledge-intensive tasks has consistently proven beneficial to downstream performance. Nonetheless, current integration techniques require extensive retraining or finetuning. To overcome this issue, we introduce KoRe, a methodology to encode 1-hop sub-graphs into compact discrete knowledge tokens and inject them into a LLM backbone. We test the proposed approach on three established benchmarks, and report competitive performances coupled with a significant reduction (up to 10x) in token usage. Our results show that compact discrete KG representations can efficiently and effectively be used to ground modern LLMs.
From Structured Knowledge to More Trustworthy AI
The research presented at GenAIK–NORA reflects a broader challenge at the heart of current AI development: how can increasingly powerful generative systems become more reliable, interpretable, and grounded in knowledge that can be explicitly represented and managed?
The GenAIK–NORA workshop brought together researchers working across Generative AI, Large Language Models, Knowledge Graphs, Semantic Web technologies, and agentic systems, with particular attention to knowledge grounding, explainability, reasoning, and trustworthy AI.
These themes strongly resonate with the DataScientia vision. DataScientia promotes research that transforms data and community knowledge into trusted knowledge capable of supporting inclusive, diversity-aware, locally grounded, and human-centred AI systems. Its research activities emphasize collaboration between people, communities, universities, and researchers in building more responsible AI futures. Work such as KoRe contributes to this direction by exploring new ways of connecting structured knowledge with generative AI—an important step toward AI systems whose knowledge can be more explicitly represented, grounded, and potentially updated and examined.
Celebrating Our Community
At DataScientia, we believe that research excellence grows through collaboration and an active community in which people learn, contribute, and advance knowledge together. Recognition such as the Best Student Paper Award highlights not only an individual achievement, but also the value of supporting emerging researchers and innovative ideas that can shape the future of AI.
We warmly congratulate Davide Cavicchini, Fausto Giunchiglia, and Jacopo Staiano on this achievement and look forward to seeing how their work on knowledge-grounded language models develops further.
- Read the paper: KoRe: Compact Knowledge Representations for Large Language Models
- Explore the project code: KoRe on GitHub
- Learn more about the GenAIK 2026 Workshop: Generative AI and Knowledge Graphs (GenAIK)
- Learn more about DataScientia Research: DataScientia Research Activities
Keywords: Artificial Intelligence, Generative AI, Large Language Models, LLMs, Knowledge Graphs, Knowledge Representation, Trustworthy AI, AI Research, Semantic Web, AI and Knowledge Graphs, KoRe, GenAIK 2026, NORA 2026, IJCAI-ECAI 2026, Best Student Paper Award
Tag:knowledge graphs, LLMs, research, success
