UNIST AIGS/IE Service Engineering & Knowledge Discovery Lab (Prof. Chiehyeon Lim)’s paper has been accepted to the Neural Information Processing Systems (NeurIPS) Datasets and Benchmarks Track 2021:
“MIND dataset for diet planning and dietary healthcare with machine learning: Dataset creation using combinatorial optimization and controllable generation with domain experts” by Changhun Lee*, Soohyeok Kim*, Sehwa Jeong, Chiehyeon Lim†, Jayun Kim, Yeji Kim, Minyoung Jung†
Diet planning, a basic and regular human task, is important to all individuals, from children to seniors and from healthy people to patients. Many recent attempts have been made to develop machine learning (ML) applications related to diet planning and dietary health research. However, given the complexity and difficulty of this task, no high-quality diet-level dataset exists at present, even among professionals, such as dietitians and physicians. In this work, we define a diet as the sequence of menus and create the Menus–Ingredients–Nutrients–Diets (MIND) dataset for the ML tasks regarding diet planning and dietary health research. The MIND dataset was created by integrating the capabilities of an operations research (OR) model that specifies and applies explicit data requirements for diet solution generation, experts who can consider implicit data requirements to make diet sequences realistic, and a controllable generation machine that automates the high-quality diet generation process. The MIND dataset can be easily downloaded and used with the Python package dietkit, which is accessible via the package installer for Python. This work is expected to contribute to the use of ML in solving medical, economic, and social problems associated with diet planning.