Year: 2025
Type: book
Abstract: Despite the potential of Knowledge Graphs (KGs) in enhancing treatment recommender systems (RSs), there is a significant research gap in developing ontologies and KGs that effectively capture ethical, legal, and societal aspects (ELSA) in healthcare RSs. This study explores the use of Large Language Models (LLMs) to assist in generating KGs that incorporate ELSA for a dairy-cow treatment RS. Our approach involves the generation of five KGs: for cows (the users); treatments (the items); ethical considerations; legal requirements, and social preferences. The KG generation process is semi-automatic, with LLMs aiding in generating structured tuples based on predefined ontologies. Researchers manually curate the information to ensure accuracy and relevance. Our study focuses on the ontological modeling of the KGs, demonstrating how entities and relationships are defined to capture the complex interplay between veterinary treatments and ELSA compliance. Use-inspired scenarios illustrate the generation of KGs based on the ontologies tailored to Dutch dairy production. The resulting KGs provide a foundation for developing dairy treatment RSs that consider ELSA. Our results contribute to integrating domain expertise with AI tools to create knowledge structures supporting informed and ELSA-compliant livestock management decision-making.
Authors: Kalidas, Varsha, Regino, André Gomes, Rossanez, Anderson, dos Reis, Julio Cesar, Alskaif, Tarek, da Silva Torres, Ricardo
Institution: Universidade Estadual de Campinas (UNICAMP)
Language: English
Topic: Semantic Web and Ontologies
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7110653307