Designing and Evaluating a Nutrition Recommender System for Improving Food Security in a Developing Country Publisher Pubmed



Abhari S^{1, 2} ; Lankarani KB³ ; Azadbakht L⁴ ; Niakan Kalhori SR^{1, 5} ; Safdari R¹ ; Sefiddashti SE⁶ ; Garavand A⁷ ; Barzegari S⁸ ; Moradi S⁹
Source: Archives of Iranian Medicine Published:2023

Abstract

Background: Due to the increased price of foods in recent years and the diminished food security in Iran, nutrition recommender systems can suggest the most suitable and affordable foods and diets to users based on their health status and food preferences. Objective: The present study aimed to design and evaluate a recommender system to suggest healthy and affordable meals and provide a tele-nutrition consulting service. Methods: This applied three-phase study was conducted in 2020. In the first stage, the food items’ daily prices were extracted from credible sources, and accordingly, meals were placed in three price categories. After conducting a systematic review of similar systems, the requirements and data elements were specified and confirmed by 10 nutritionists and 10 health information management and medical informatics experts. In the second phase, the software was designed and developed based on the findings. In the third phase, system usability was evaluated by four experts based on Nielsen’s heuristic evaluation. Results: Initially, 72 meals complying with nutritional principles were placed in three price categories. Following a literature review and expert survey, 31 data elements were specified for the system, and the experts confirmed system requirements. Based on the information collected in the previous stage, the Web-based software TanSa in the Persian language was designed, developed, and presented on a unique domain. During the evaluation, the mean severity of the problems associated with Nielsen’s 10 principles was 1.2, which is regarded as minor. Conclusion: To promote food security, the designed system recommends healthy, nutritional, and affordable meals to individuals and households based on user characteristics. © 2023 Academy of Medical Sciences of I.R. Iran. All rights reserved.