JMIR Publications

ABSTRACT

Background:

Globally, the number of people suffering from chronic diseases such as heart disease, stroke, diabetes, and hypertension is increasing. Healthy eating may assist people with chronic diseases to manage their conditions and avoid complications. However, it is not easy for people to make healthy nutritious meal plans, as different people may have different health concerns, nutritional requirements, tastes, economic status, time limits, etc. There is an urgent need for effective, affordable, and healthy meal planning that considers various contexts and factors that affect people’s choice of food.

Objective:

The goal of this study is to design an AI-empowered meal planner that can assist people to make personalized healthy meal plans based on their specific health conditions, personal preferences, and status.

Methods:

We propose a system that integrates semantic reasoning, fuzzy logic, heuristic search, and multi-criteria analysis to produce flexible optimized meal plans based on the user’s health concerns, nutrition needs, food restrictions/constraints, and all other personal references. Specifically, an ontology-based knowledge base is constructed to model knowledge about food and nutrition. Semantic rules are defined to represent diet guidelines for different health concerns. Fuzzy membership of food nutrition is built on the experience of experts to handle nutrition data that is vague and lacks certainty, so we can use non-linear functions to model nutrition. We applied a semantic rule-based filtering mechanism to filter out food that violates mandatory health guidelines and constraints such as allergies and religion. We designed a novel heuristic search method that identifies the finest meals among hundreds of thousands of candidates and evaluates them based on their fuzzy nutritional score. To select nutritious meals that also satisfy the user’s other preferences, a multi-criteria decision-making approach was proposed.

Results:

A mobile application prototype system was implemented. Use case study and user study of the application were performed to evaluate its effectiveness. The use case study and the user study based on the prototype demonstrate that the proposed system has the following advantages: 1. A user’s health concerns have been well considered. 2. Nutrition values are optimized. 3. Dietary, religious, and other constraints and restrictions are fully respected. 4. User preferences are considered. 5. Users are satisfied with the system in general.

Conclusions:

We designed an AI-powered meal planner that helps people make healthy and personalized meal plans based on their health conditions, preferences, and status. Our use case study and user study confirmed the usability and feasibility of the proposed system.