Case-Based Reasoning and Computational Creativity in a Recipe Recommender System

Abstract

A domain where humans have unfolded their creativity for thousands of years is cooking. However, can a human's creativity within cooking be transferred to a computer program? The case-based reasoning (CBR) methodology allows to incorporate creativity when using earlier experiences to solve new problems. The main objective of our research is to design and construct a CBR based recipe recommender system that enhances creativity to adapt recipes based on a given user query.

We have reviewed related work performed in the field of recipe recommendation and identified typical approaches and features within those. Further, we developed a knowledge engineering heavy system that retrieve, compare, adapt, and suggest recipes given a user query containing desired and undesired ingredients.

The system was tested with controlled observations, questionnaires, and an online quiz. Evaluation results show that the system appears user-friendly, self-explanatory, and with meaningful recipe recommendations. The resulting system is able to adapt recipes in a way that humans find it challenging to distinguish them from human created recipes.