Abstract
OBJECTIVE: To develop a research agenda to investigate the effectiveness of AI-tailored advice to support healthier home cooking. It aims to support healthier food choice in the context of hypertension, allergies, and sustainable diets. METHODS: We describe an agenda that has been formed between 2019 and 2022, through multiple rejected grant applications to the Research Council of Norway. We focus on the case of tailored recipe advice for individuals, formulating research questions and methods for three topics: "Acceptance of Personalized Food Advice", "Algorithm and Interface AI: App Development", and "Nutrition Modeling & Clinical Trials". The overall methodology focuses on mitigating health issues among individuals with hypertension. CONCLUSION: The design of AI to support healthier home cooking should tap into computational principles, as well as (psychological) theories of behavioral change. The effectiveness of an AI-driven home cooking app can be evaluated in a clinical trial akin to 'regular' dietary intervention studies. INNOVATION: The development of a research agenda requires an integrated effort between scientists from different domains, during both the development and writeup of ideas. The proposed project is innovative, as most food technology and AI approaches have yet to be tested in proper trials on changes in eating habits.