FoodKG

FoodKG is a pioneering software tool that enhances the utility and accessibility of datasets in the Food, Energy, and Water (FEW) domains by leveraging advanced machine learning techniques to enrich FEW knowledge graphs. This tool addresses the critical need for reliable methods to utilize the abundant but underutilized FEW datasets available on the Internet, aiming to bolster decision-making, knowledge discovery, and improve search functionalities for data scientists working within these sectors.

Utilizing an input knowledge graph built from raw FEW datasets, FoodKG enriches it by adding semantically related triples, relations, and images, drawing upon terms and classes from the original dataset. It employs graph embedding techniques trained on AGROVOC, a comprehensive controlled vocabulary curated by the Food and Agriculture Organization of the United Nations, which encompasses a wide range of terms and classes pertinent to agriculture and food. Through this approach, FoodKG can augment knowledge graphs with semantic similarity scores, establish relations between different classes, classify existing entities, and facilitate the use of scientific terms for articulating FEW concepts accurately.