Characterizing and Predicting Repeat Food Consumption Behavior for Just-in-Time Interventions

Abstract

Human beings are creatures of habit. In their daily life, people tend to repeatedly consume similar types of food items over several days and occasionally switch to consuming different types of items when the consumptions become overly monotonous. However, the novel and repeat consumption behaviors have not been studied in food recommendation research. More importantly, the ability to predict daily eating habits of individuals is crucial to improve the effectiveness of food recommender systems in facilitating healthy lifestyle change. In this study, we analyze the patterns of repeat food consumptions using large-scale consumption data from a popular online fitness community called MyFitnessPal (MFP), conduct an offline evaluation of various state-of-the-art algorithms in predicting the next-day food consumption, and analyze their performance across different demographic groups and contexts. The experiment results show that algorithms incorporating the exploration-andexploitation and temporal dynamics are more effective in the nextday recommendation task than most state-of-the-art algorithms.

Effects and challenges of using a nutrition assistance system: results of a long-term mixed-method study

“…The dependency on well-kept food diaries, albeit theoretically achieving high degrees of personalization and accuracy, is arguably the main drawback of Nutrilize's recommendation feature. Future systems should consider implementing temporal and intra-list diversity metrics for long-term usage as well as approximating missing dietary diary information by default or substitute learning (Liu et al 2019).…”

Food Habits: Insights from Food Diaries via Computational Recurrence Measures

“…Consumption, broadly defined across many behaviors, is often characterized as a combination of novel (exploration) and repetitive events (exploitation) [ 2 ]. Similarly, in the domain of food consumption, we can characterize human behavior associated with food choice selection as a mixture of novelty-seeking choices and habitual choices [ 3 , 4 , 5 ]. The novelty-seeking consumption corresponds to new or occasional food items consumed by an individual.…”

“…Smartphone-based applications [ 13 ] for daily food logging are also emerging as a promising instrument for understanding food consumption behaviors [ 14 ]. In recent works [ 5 , 15 ], the authors utilized food diaries to study and model food consumption behavior. Related important past works [ 5 ] studied the repeat consumption of food choices and created prediction models to forecast the top-k food choices each day for users of a food diary dataset [ 16 ].…”

“…In recent works [ 5 , 15 ], the authors utilized food diaries to study and model food consumption behavior. Related important past works [ 5 ] studied the repeat consumption of food choices and created prediction models to forecast the top-k food choices each day for users of a food diary dataset [ 16 ]. Another insightful study [ 15 ] used the daily logged data of individuals to recommend easy to adopt personalized food suggestions for behavior change by leveraging frequent past healthy behaviors and occasionally exploring infrequent behaviors, while past works [ 5 , 15 ] have demonstrated the utility of habitual food consumption behavior.…”

“…Characteristics of habitual consumption: Previous work [ 5 ] discussed repeat consumption in the context of habitual consumption and defined repeat consumption behavior as the average fraction of items consumed in a day that was consumed in the last seven days. Since repeat consumption captures short-term repetitions, it is a short-term characterization of habitual consumption.…”