Digital dietary coaches can supplement current approaches for guiding consumers towards healthier behavior. In addition to taking into account the individual client’s health status, digital coaches must also link to her or his personal preferences and habits and to any contextual factors such as location and time of the day. We address the question which food attributes are needed to generate an advice that is fully personalized and situational. In the first part of this study, we have made a systematic analysis of food item attributes and the way in which they are used in making food choices. We distinguish between (1) food attributes as such, (2) consumer attributes (preferences, profile, habits, etc.) and (3) context attributes(time, occasion, situation, etc.). One source for finding attributes is human behavior theory. In addition, we apply a more empirical approach by analyzing attributes that dietitians and food experts use in practice. In the second part of this study we have asked consumers to indicate which food selection criteria they use in practice. We list the sets of attributes and discuss how they can be used to automatically infer preferred food items from personal and situational data. Although fully automatic generation of food alternatives is not yet possible, dietitians and life style coaches can benefit from the data that is already available. A number of web services is currently under development to access the data programmatically, for example in dietary apps.
|Name||Report / Wageningen Food & Biobased Research|