Intelligent Computing Approaches for Nutrition-aware Food Recommendation System: Concepts, Techniques and Trends

Maintaining a healthy lifestyle in people's lives needs recurring and non-trivial expert's help and time commitment as per the expert's schedule, in addition to the expenses. The computerized solution is known under various terms, such as recipe, nutrition, and ingredient recommendation, either for specific diseases or for general use. These solutions are either mobile or web applications, available for free or at very low cost. They integrate various computing techniques for recommendation, including optimization, meta-heuristics, case-based reasoning, knowledge-based, and learning-based approaches. Following an extensive review of the nutrition literature, we felt the need for a systematic survey of these existing solutions, which is discussed in this paper using a multi-level taxonomy. At the first level, we have considered 5 categories. Further, they have been categorized at the second level. Each category is compared across various parameters. The current issues/challenges in food/recipe recommendation are highlighted with future directions.