Ontological how and why: action and objective of planned processes in the food domain

The computational modeling of food processing, aimed at various applications including industrial automation, robotics, food safety, preservation, energy conservation, and recipe nutrition estimation, has been ongoing for decades within food science research labs, industry, and regulatory agencies. The datasets from this prior work have the potential to advance the field of data-driven modeling if they can be harmonized, but this requires a standardized language as a starting point. Our primary goal is to explore two interdependent aspects of this language: the granularity of process modeling sub-parts and parameter details and the substitution of compatible inputs and processes. A delicate semantic distinction-categorizing planned processes based on the objectives they seek to fulfill vs. categorizing them by the actions or mechanisms they utilize-helps organize and facilitate this endeavor. To bring an ontological lens to process modeling, we employ the Open Biological and Biomedical Ontology Foundry ontological framework to organize two main classes of the FoodOn upper-level material processing hierarchy according to objective and mechanism, respectively. We include examples of material processing by mechanism, ranging from abstract ones such as "application of energy" down to specific classes such as "heating by microwave." Similarly, material processing by objective-often a transformation to bring about materials with certain qualities or composition-can, for example, range from "material processing by heating threshold" to "steaming rice".

Machine learning models to predict micronutrient profile in food after processing

The information on nutritional profile of cooked foods is important to both food manufacturers and consumers, and a major challenge to obtaining precise information is the inherent variation in composition across biological samples of any given raw ingredient. The ideal solution would address precision and generability, but the current solutions are limited in their capabilities; analytical methods are too costly to scale, retention-factor based methods are scalable but approximate, and kinetic models are bespoke to a food and nutrient. We provide an alternate solution that predicts the micronutrient profile in cooked food from the raw food composition, and for multiple foods. The prediction model is trained on an existing food composition dataset and has a 31% lower error on average (across all foods, processes and nutrients) than predictions obtained using the baseline method of retention-factors. Our results argue that data scaling and transformation prior to training the models is important to mitigate any yield bias. This study demonstrates the potential of machine learning methods over current solutions, and additionally provides guidance for the future generation of food composition data, specifically for sampling approach, data quality checks, and data representation standards.

Using Word Embeddings to Learn a Better Food Ontology

Food ontologies require significant effort to create and maintain as they involve manual and time-consuming tasks, often with limited alignment to the underlying food science knowledge. We propose a semi-supervised framework for the automated ontology population from an existing ontology scaffold by using word embeddings. Having applied this on the domain of food and subsequent evaluation against an expert-curated ontology, FoodOn, we observe that the food word embeddings capture the latent relationships and characteristics of foods. The resulting ontology, which utilizes word embeddings trained from the Wikipedia corpus, has an improvement of 89.7% in precision when compared to the expert-curated ontology FoodOn (0.34 vs. 0.18, respectively, p value = 2.6 × 10^–138), and it has a 43.6% shorter path distance (hops) between predicted and actual food instances (2.91 vs. 5.16, respectively, p value = 4.7 × 10^–84) when compared to other methods. This work demonstrates how high-dimensional representations of food can be used to populate ontologies and paves the way for learning ontologies that integrate contextual information from a variety of sources and types.