Aligning Food Ingredients with Multiple Semantic Resources
Authors
D. Sasanski, A. Todorovski, B. Trpeski, D. Trajanov, T. Eftimov, R. Stojanov
Publication
ICT Innovations 2024 ICT Innovations 2024
Skopje, North Macedonia, April 23, 2025
Abstract
To address the lack of integrated resources that combine nutritional, chemical, and semantic information, we have integrated the Recipe1M+ dataset with the USDA National Nutrient Database, FooDB, and the FoodOn ontology. This integration brings significant advantages by enriching the dataset with comprehensive information across multiple domains. By linking Recipe1M+ ingredients to the USDA database, we provide detailed nutritional profiles that enable a comprehensive analysis of recipes, supporting researchers and dietitians in developing personalized nutrition plans based on accurate nutrient data. Furthermore, incorporating FooDB enhances the dataset with in-depth chemical compositions and health effects of food constituents, facilitating research on functional foods and their role in disease prevention and health promotion. Mapping ingredients to FoodOn further expands the dataset’s semantic context, encompassing food products, production, agriculture, and environmental impacts.
BIBTEX copied to Clipboard