Datasets
Standard Dataset
CRWB: A Multilingual Ontology and Knowledge Graph for Recipes
- Citation Author(s):
-
MansiGOEL
IIITDFredericANDRESNII - Submitted by:
- Andres Frederic
- Last updated:
- Mon, 03/03/2025 - 02:17
- DOI:
- 10.21227/pngr-dr67
- Data Format:
- License:
50 Views- Categories:
- Keywords:
Abstract
In recent years, the fusion of artificial intelligence and semantic web technologies has paved the way for innovative approaches to managing and utilizing information. With the growing demand for structured gastronomical data, there is a need for well-defined ontologies that facilitate recipe organization, ingredient classification, nutritional insights, and personalized diet recommendations. The dataset presents a multilingual recipe ontology and knowledge graph, capturing critical relationships between ingredients, nutrition, cooking actions, and recipe planning. Our ontology supports interoperability across different languages (English, Hindi, and Japanese), facilitating AI applications such as ingredient substitution and personalized recommendations. The proposed knowledge graph leverages semantic web technologies to enhance structured data accessibility and machine-readable representations, ultimately contributing to computational gastronomy and food informatics.
Files in the Project (zip)
1. MG_Ontology_Modified_21Feb.ttl
- This file contains the multilingual ontology in Turtle and OWL format.
- It includes structured representations of recipes, ingredients, nutrition information, ingredient-nutrition relationships, and cooking actions.
2. Metrics_with_KnownURI.ipynb
- This Jupyter Notebook provides ontology quality metrics for cases where the URIs of entities are known.
- It helps visualize and analyze the ontology structure using predefined URIs.
3. Quality_Metrics.ipynb
- This Jupyter Notebook provides generalized quality metrics to evaluate the ontology.
- It does not require predefined URIs, making it more flexible for evaluating various ontology versions.
How to Use
1. Ontology Evaluation
- Open 'Quality_Metrics.ipynb' to evaluate the ontology quality without requiring known URIs.
- If you have predefined URIs, use 'Metrics_with_KnownURI.ipynb' for more detailed analysis.
2. Ontology Visualization
- You can use tools like Protégé or Neo4j with RDF plugins to visualize the ontology from 'MG_Ontology_Modified_21Feb.ttl'.
Requirements
- Python 3.x
- Jupyter Notebook
- RDFLib, Owlready2
- Neo4j with RDF plugin or WebVOWL for visualization
