Algorithms and data-models for modern context-aware nutrition-advice applications (FA5 module2)

Nowadays, there is a wide variety of mobile applications to support healthy living and a healthy diet. Nevertheless, many users struggle to incorporate the provided advice into their daily lives because the recommendations often conflict with individual limitations, needs and situations.

The topic of Focus Area 5 is therefore a virtual dietary advisor, which aims at catering to users’ individual needs much more. First approaches for personalized nutrition recommendations have been investigated in the first funding period of enable and will now be extended upon.Module 2 of Focus Area 5 deals with algorithms and data-models for the virtual dietary advisor, i.e. the background structures and procedures that are needed to provide personalized, context-aware and anticipatory dietary advice.

The virtual dietary advisor is supposed to help the users in situations when decisions about buying or consuming food have to be made. Furthermore, nutrition recommendations provided by the application should be seen as feasible alternatives in order to increase practicability and through this the acceptance of dietary recommendations. Therefore, a close intertwinement between dietary advice and a user’s individual situation is aimed for. This includes an expressive user profile, which for instance reflects personal culinary preferences. Additionally, the user’s current context needs to be considered, enabling the virtual advisor to act and react accordingly. As an example, the issue of overeating under stress can be addressed if the application can take a user’s emotional state and situation into account.

Our objectives

The goal of this project is the development of algorithms and data-models which allow a virtual dietary advisor to give personalized, context-aware and anticipatory dietary advice. This includes recommending healthy food-substitutes which correspond to the original food item’s culinary attributes such as taste and texture, as well as recommending eating and shopping plans that flexibly adapt to users and their current contexts.