Elamur Research⌕
☰
home / papers / 2026 / Thesis Proposal: A Multi-Agent System for Ontology-Based Perspective-Aware Knowledge Extraction

Thesis Proposal: A Multi-Agent System for Ontology-Based Perspective-Aware Knowledge Extraction

Paper · 2026 · Luiz Do Valle Miranda and Grzegorz J. Nalepa

Summary

Most knowledge engineering and LLM pipelines merge differing viewpoints into one consensus schema. This hides where claims come from and erases legitimate disagreement. This thesis proposal from Jagiellonian University plans to keep competing stakeholder views as separate, queryable layers within a knowledge graph (KG). It combines two lines of work: polyvocal ontologies, which so far appear mostly in cultural heritage, and LLM-based multi-agent systems (MAS).

Three research questions structure the project:

  • using LLMs to generate polyvocal Ontology Design Patterns (ODPs), including retrofitting univocal ontologies such as SNOMED CT;
  • designing a MAS in which each agent extracts facts through a perspective-specific ontology module while preserving schema coherence and provenance;
  • testing whether the extracted facts are genuinely diverse without damaging KG integrity.

The target domain is medicine. Data will come from the Jagiellonian University Medical Faculty, supplemented by public sources. No empirical results are reported yet. The work matters because it would extend NLP perspectivism beyond annotation and evaluation into knowledge-based systems for high-stakes fields.

Key points

  • The three RQs cover polyvocal ODP design (RQ1), ontology-grounded MAS extraction (RQ2), and evaluation of semantic diversity against schema coherence and inter-agent consistency (RQ3).
  • The project aims to deliver a library of medical polyvocal ODPs, a fact-extraction framework whose provenance can be audited, and an empirical evaluation conducted together with domain experts.
  • MIMIC-III is being assessed as a public dataset. Wikidata qualifiers and DBpedia's multilingual representations are named as inspiration for ODP design.
  • If LLM-generated ODPs are poor, the fallback is human-in-the-loop design. Experts would define perspective distinctions in workshops, LLMs would refine them, and OWL reasoners plus SPARQL competency questions would check coherence.
  • Private patient data cannot be shared, so the authors plan to release open-source code and evaluation scripts and possibly a synthetic benchmark. They will run only locally hosted models and will seek ethics committee approval.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Jagiellonian UniversityKrakow · Poland4 works · 256 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksBiochemistry, Genetics and Molecular Biology · 234 worksMolecular Biology · 231 works

🧭 Topics 1Research subjects and languages the work is about.Knowledge graphs & ontologies · 192 works

Cite

Cite as Luiz Do Valle Miranda and Grzegorz J. Nalepa. 2026. Thesis Proposal: A Multi-Agent System for Ontology-Based Perspective-Aware Knowledge Extraction. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 4: Student Research Workshop).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.eacl-srw.46

@inproceedings{miranda2026thesis,
  title = {{Thesis Proposal: A Multi-Agent System for Ontology-Based Perspective-Aware Knowledge Extraction}},
  author = {Luiz Do Valle Miranda and Grzegorz J. Nalepa},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 4: Student Research Workshop)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-srw.46},
  url = {https://doi.org/10.18653/v1/2026.eacl-srw.46}
}

References

21 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency, FAccT ’24, page 1433–1445, New York, NY, USA. Association for Computing Machinery. find
  2. Federico Cabitza, Andrea Campagner, and Valerio Basile. 2023. Toward a perspectivist turn in ground truthing for predictive computing. Proceedings of the AAAI Conference on Artificial Intelligence, 37(6):6860–6868.
  3. Simona Frenda, Gavin Abercrombie, Valerio Basile, Alessandro Pedrani, Raffaella Panizzon, Alessandra Teresa Cignarella, Cristina Marco, and Davide Bernardi. 2024. Perspectivist approaches to natural language processing: a survey: Perspectivist approaches to natural language processing... Lang. Resou… find
  4. Isabel O. Gallegos, Ryan A. Rossi, Joe Barrow, Md Mehrab Tanjim, Sungchul Kim, Franck Dernoncourt, Tong Yu, Ruiyi Zhang, and Nesreen K. Ahmed. 2024. Bias and fairness in large language models: A survey. Computational Linguistics, 50(3):1097– 1179.
  5. Aldo Gangemi and Valentina Presutti. 2022. Formal representation and extraction of perspectives. In Piek Vossen and Antske Fokkens, editors, Creating a More Transparent Internet, pages 208–228. Cambridge University Press, Cambridge. find
  6. Anna Sofia Lippolis, Mohammad Javad Saeedizade, Robin Keskisärkkä, Sara Zuppiroli, Miguel Ceriani, Aldo Gangemi, Eva Blomqvist, and Andrea Giovanni Nuzzolese. 2025. Ontology generation using large language models. In The Semantic Web: 22nd European Semantic Web Conference, ESWC 2025, Portoroz, Slove… find
  7. Sanaz Saki Norouzi, Adrita Barua, Antrea Christou, Nikita Gautam, Andrew Eells, Pascal Hitzler, and Cogan Shimizu. 2024. Ontology population using llms. Preprint, arXiv:2411.01612. — [arXiv:2411.01612](https://arxiv.org/abs/2411.01612) find
  8. Sanghyun Park, Boris Maciejovsky, and Phanish Puranam. 2025. Thinking with many minds: Using large language models for multi-perspective problem-solving. Preprint, arXiv:2501.02348. — [arXiv:2501.02348](https://arxiv.org/abs/2501.02348) find
  9. Sami Saadaoui and Eduardo Alonso. 2025. Coordinated llm multi-agent systems for collaborative question-answer generation. Knowledge-Based Systems, 330:114627. find
  10. Alireza Ghafarollahi and Markus J. Buehler. 2025. Sciagents: Automating scientific discovery through bioinspired multi-agent intelligent graph reasoning. Advanced Materials, 37(22):2413523. find
  11. Cogan Shimizu and Pascal Hitzler. 2025. Accelerating knowledge graph and ontology engineering with large language models. Journal of Web Semantics, 85:100862. find
  12. Shirley Anugrah Hayati, Minhwa Lee, Dheeraj Rajagopal, and Dongyeop Kang. 2024. How far can we extract diverse perspectives from large language models? In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 5336–5366, Miami, Florida, USA. Association for Com… find
  13. Sarah Binta Alam Shoilee, Victor de Boer, and Jacco van Ossenbruggen. 2023. Polyvocal Knowledge Modelling for Ethnographic Heritage Object Provenance, pages 127–143. Studies on semantic web. IOS Press. find
  14. Alistair Johnson, Tom Pollard, and Roger Mark. 2016. MIMIC-III clinical database (version 1.4). PhysioNet. find
  15. Jackie Kay, Atoosa Kasirzadeh, and Shakir Mohamed. 2025. Epistemic Injustice in Generative AI, page 684–697. AAAI Press. find
  16. Samira Khorshidi, Azadeh Nikfarjam, Suprita Shankar, Yisi Sang, Yash Govind, Hyun Jang, Ali Kasgari, Alexis McClimans, Mohamed Soliman, Vishnu Vardhan Reddy Konda, Ahmed Fakhry, and Xiaoguang Qi. 2025. Odke+: Ontology-guided open-domain knowledge extraction with llms. ArXiv, abs/2509.04696. — [arXiv:2509.04696](https://arxiv.org/abs/2509.04696) find
  17. Grgur Kovač, Masataka Sawayama, Rémy Portelas, Cédric Colas, Peter Ford Dominey, and Pierre-Yves Oudeyer. 2023. Large language models as superpositions of cultural perspectives. ArXiv, abs/2307.07870. — [arXiv:2307.07870](https://arxiv.org/abs/2307.07870) find
  18. Marieke van Erp and Victor de Boer. 2021. A polyvocal and contextualised semantic web. In The Semantic Web, pages 506–512, Cham. Springer International Publishing. find
  19. Bohui Zhang, Valentina Anita Carriero, Katrin Schreiberhuber, Stefani Tsaneva, Lucía Sánchez González, Jongmo Kim, and Jacopo de Berardinis. 2025. Ontochat: A framework for conversational ontology engineering using language models. In The Semantic Web: ESWC 2024 Satellite Events, pages 102–121, Cham… find
  20. Kaiwen Zuo, Yirui Jiang, Fan Mo, and Pietro Lio. 2024. Kg4diagnosis: A hierarchical multi-agent llm framework with knowledge graph enhancement for medical diagnosis. ArXiv, abs/2412.16833. — [arXiv:2412.16833](https://arxiv.org/abs/2412.16833) find
  21. Angelie Kraft and Eloïse Soulier. 2024. Knowledge-enhanced language models are not bias-proof: Situated knowledge and epistemic injustice in ai. In find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/miranda-2026-thesis-proposal-multi-agent-system/