Publications
2026
Grewenig, Lisa; Zeyen, Christian; Schultheis, Alexander; Malburg, Lukas; Bergmann, Ralph
Challenges and Support Potentials in the Development of CBR Applications Proceedings Article
In: Case-Based Reasoning Research and Development - 34th International Conference, ICCBR 2026, Bremen, Germany, August 13 - 16, 2026, Proceedings, Springer, 2026, (Accepted for Publication.).
Abstract | Links | BibTeX | Tags: Case-Based Reasoning, CBR Applications, Development Support, Domain Experts, Knowledge Engineering
@inproceedings{GrewenigZSMB2026,
title = {Challenges and Support Potentials in the Development of CBR Applications},
author = {Lisa Grewenig and Christian Zeyen and Alexander Schultheis and Lukas Malburg and Ralph Bergmann},
url = {https://www.wi2.uni-trier.de/shared/publications/2026_ICCBR_GrewenigEtAl.pdf},
doi = {10.1007/978-3-032-33865-5_36},
year = {2026},
date = {2026-01-01},
booktitle = {Case-Based Reasoning Research and Development - 34th International Conference, ICCBR 2026, Bremen, Germany, August 13 - 16, 2026, Proceedings},
volume = {16780},
publisher = {Springer},
series = {Lecture Notes in Computer Science},
abstract = {Developing Case-Based Reasoning applications is a knowledge-intensive task that consists of numerous demanding and time-consuming development steps. Major challenges are posed by knowledge engineering since CBR systems depend on capturing, validating, and maintaining domain knowledge in an explicit, machine-interpretable form. This usually can be done only in a close interdisciplinary collaboration between CBR and domain experts. To support development, previous works put forth a broad spectrum of support functions but mostly have not yet addressed the synergies of support functions or the potentials for involving domain experts, particularly those without knowledge-engineering experience. To further improve the accessibility of CBR frameworks and to reduce the effort of development, this paper consolidates development challenges along with corresponding support methods. In addition, we address the potentials for involving domain experts in the development with the help of support functions. We derived development challenges and support potentials from literature and conducted an explorative survey in the CBR research community.},
note = {Accepted for Publication.},
keywords = {Case-Based Reasoning, CBR Applications, Development Support, Domain Experts, Knowledge Engineering},
pubstate = {published},
tppubtype = {inproceedings}
}
Developing Case-Based Reasoning applications is a knowledge-intensive task that consists of numerous demanding and time-consuming development steps. Major challenges are posed by knowledge engineering since CBR systems depend on capturing, validating, and maintaining domain knowledge in an explicit, machine-interpretable form. This usually can be done only in a close interdisciplinary collaboration between CBR and domain experts. To support development, previous works put forth a broad spectrum of support functions but mostly have not yet addressed the synergies of support functions or the potentials for involving domain experts, particularly those without knowledge-engineering experience. To further improve the accessibility of CBR frameworks and to reduce the effort of development, this paper consolidates development challenges along with corresponding support methods. In addition, we address the potentials for involving domain experts in the development with the help of support functions. We derived development challenges and support potentials from literature and conducted an explorative survey in the CBR research community.
2024
Brand, Florian; Malburg, Lukas; Bergmann, Ralph
Large Language Models as Knowledge Engineers Proceedings Article
In: Malburg, Lukas (Ed.): Proceedings of the Workshops at the 32nd International Conference on Case-Based Reasoning (ICCBR-WS 2024) co-located with the 32nd International Conference on Case-Based Reasoning (ICCBR 2024), Mérida, Mexico, July 1, 2024, pp. 3–18, CEUR-WS.org., 2024.
Abstract | Links | BibTeX | Tags: {Case-Based Reasoning, Knowledge Acquisition Bottleneck, Knowledge Engineering, Large Language Models, Prompting}
@inproceedings{Brand.2024_LLMKnowledgeEngineer,
title = {Large Language Models as Knowledge Engineers},
author = {Florian Brand and Lukas Malburg and Ralph Bergmann},
editor = {Lukas Malburg},
url = {https://www.wi2.uni-trier.de/shared/publications/2024_ICCBR-WS_LLMInCBR_BrandEtAl.pdf},
year = {2024},
date = {2024-01-01},
booktitle = {Proceedings of the Workshops at the 32nd International Conference
on Case-Based Reasoning (ICCBR-WS 2024) co-located with the 32nd
International Conference on Case-Based Reasoning (ICCBR 2024), Mérida,
Mexico, July 1, 2024},
volume = {3708},
pages = {3–18},
publisher = {CEUR-WS.org.},
series = {CEUR Workshop Proceedings},
abstract = {Many Artificial Intelligence (AI) systems require human-engineered knowledge at their core to reason about new problems based on this knowledge, with Case-Based Reasoning (CBR) being no exception. However, the acquisition of this knowledge is a time-consuming and laborious task for the domain experts that provide the needed knowledge. We propose an approach to help in the creation of this knowledge by leveraging Large Language Models (LLMs) in conjunction with existing knowledge to create the vocabulary and case base for a complex real-world domain. We find that LLMs are capable of generating knowledge, with results improving by using natural language and instructions. Furthermore, permissively licensed models like CodeLlama and Mixtral perform similar or better than closed state-of-the-art models like GPT-3.5 Turbo and GPT-4 Turbo.},
keywords = {{Case-Based Reasoning, Knowledge Acquisition Bottleneck, Knowledge Engineering, Large Language Models, Prompting}},
pubstate = {published},
tppubtype = {inproceedings}
}
Many Artificial Intelligence (AI) systems require human-engineered knowledge at their core to reason about new problems based on this knowledge, with Case-Based Reasoning (CBR) being no exception. However, the acquisition of this knowledge is a time-consuming and laborious task for the domain experts that provide the needed knowledge. We propose an approach to help in the creation of this knowledge by leveraging Large Language Models (LLMs) in conjunction with existing knowledge to create the vocabulary and case base for a complex real-world domain. We find that LLMs are capable of generating knowledge, with results improving by using natural language and instructions. Furthermore, permissively licensed models like CodeLlama and Mixtral perform similar or better than closed state-of-the-art models like GPT-3.5 Turbo and GPT-4 Turbo.