Semantic Nutrition Estimation: Predicting Food Healthfulness from Text Descriptions

TL;DR

This paper introduces a machine learning system that predicts comprehensive food healthfulness scores from simple text descriptions, enabling scalable and accessible nutritional assessment without detailed data.

Contribution

It presents a novel neural network pipeline that combines semantic, lexical, and domain features to estimate nutrition scores from food descriptions, improving accessibility.

Findings

Achieved median R^2 of 0.81 for nutrient prediction

Strong correlation (r=0.77) with published Food Compass Scores

Mean absolute difference of 14 points in score predictions

Abstract

Accurate nutritional assessment is critical for public health, but existing profiling systems require detailed data often unavailable or inaccessible from colloquial text descriptions of food. This paper presents a machine learning pipeline that predicts the comprehensive Food Compass Score 2.0 (FCS) from text descriptions. Our approach uses multi-headed neural networks to process hybrid feature vectors that combine semantic text embeddings, lexical patterns, and domain heuristics, alongside USDA Food and Nutrient Database for Dietary Studies (FNDDS) data. The networks estimate the nutrient and food components necessary for the FCS algorithm. The system demonstratedstrong predictive power, achieving a median R^2 of 0.81 for individual nutrients. The predicted FCS correlated strongly with published values (Pearson's r = 0.77), with a mean absolute difference of 14.0 points. While errors were largest for ambiguous or processed foods, this methodology translates language into actionable nutritional information, enabling scalable dietary assessment for consumer applications and research.