Explanation
The chatbot is developed to provide in time personalized food suggestions. The interaction starts with understanding user preferences regarding food allergies and constraints, dish likes, and dislikes. By analyzing all those information, it proactively generates what's most relevant recipe items to the current user's food recipes in addition to a healthier alternative with a nutrition explanation.
Castiglia, Giovanni & El Majjodi, Ayoub & Calò, Federica &
Deldjoo, Yashar & Narducci, Fedelucio & Starke, Alain &
Trattner, Christoph. (2022). Nudging Towards Health in a
Conversational Food Recommender System Using Multi-Modal
Interactions and Nutrition Labels.
https://github.com/giocast/MMCFRS
Contact: PhD Candidate Ayoub el Majjodi, SFI MediaFutures, UiB: Ayoub.Majjodi@uib.no
VIDEO
Detailed Explanation
# Step One
The bot prompts you to choose your food category
# Step Two
The bot prompts you to list any preferences or food restrictions.
# Step three
The bot prompts you to list any constraints, and adjusts the recommendation thereafter.
# Step Four
Step four starts with the preference elicitation step, it asks you to like or skip at least five suggested dishes.
# Step Five
The bot delivers your personalised food recommendation with an alternative healthiest option. It then asks for user feedback.
