Transactions on Data Analysis in Social Science

Transactions on Data Analysis in Social Science

A Transparent Knowledge-Based Diet Recommender System: Evaluation on NHANES Data and a Test of the Added Value of Machine Learning

Document Type : Original Article

Authors
R. Kapoor 1
A. Roy 2
1 Department of Computer Engineering, Faculty of Engineering, University of Delhi, Delhi, India
2 Department of Computer Engineering, Jadavpur University, Kolkata, India
10.22034/tdas.2026.256211
Abstract
Overweight and obesity are major public-health problems, and access to dietary counselling is limited. This paper presents a knowledge-based diet recommender that estimates daily energy requirements from age, sex, height, weight and physical-activity level using validated equations (Mifflin–St Jeor, standard activity factors and a BMI-based adjustment with a safety floor) and recommends one of twelve dietitian-designed diets (1,000–3,500 kcal); users with diabetes or dyslipidaemia are referred to a specialist. The system was evaluated on public data from 4,785 adults in the US National Health and Nutrition Examination Survey (NHANES 2009–2012). Of these adults, 68.9% were overweight or obese and 44.6% met the referral criterion. Replacing the Mifflin–St Jeor equation with the 1918 Harris–Benedict equation left only 46.6% of non-referred adults with the same diet. An earlier prototype, which used ideal instead of actual weight and a single diet for all overweight users, assigned 77.2% of adults to one diet and agreed with the revised method in only 36.4% of the remaining cases. No machine-learning model outperformed the rule that generated the labels: the best, a multilayer perceptron, reached 92.5% accuracy, with errors clustered near energy-range boundaries. Typical self-reporting errors in weight and height changed the recommendation for 24.5% of users. A transparent, testable rule-based engine is therefore preferable to black-box models for this task; machine learning adds value only when labels come from real outcomes such as weight change or independent expert judgement. Code and unit tests are publicly available.
Keywords
Recommender system
Diet recommendation
Knowledge-based system
Basal metabolic rate
NHANES
Machine learning
Interpretability
[1]      Rajamani, K., & Sheela, D. (2018). Data mining techniques and algorithms in cloud environment. International Journal of Pure and Applied Mathematics, 119(7), 599–602.
[2]      Liang, T.-P., Lai, H.-J., & Ku, Y.-C. (2006). Personalized content recommendation and user satisfaction: Theoretical synthesis and empirical findings. Journal of Management Information Systems, 23(3), 45–70. https://doi.org/10.2753/MIS0742-1222230303
[3]      Ricci, F., Rokach, L., & Shapira, B. (2015). Recommender systems: Introduction and challenges. In F. Ricci, L. Rokach, & B. Shapira (Eds.), Recommender systems handbook (2nd ed., pp. 1–34). Springer. https://doi.org/10.1007/978-1-4899-7637-6_1
[4]      Adomavicius, G., Sankaranarayanan, R., Sen, S., & Tuzhilin, A. (2005). Incorporating contextual information in recommender systems using a multidimensional approach. ACM Transactions on Information Systems, 23(1), 103–145. https://doi.org/10.1145/1055709.1055714
[5]      Melville, P., & Sindhwani, V. (2011). Recommender systems. In C. Sammut & G. I. Webb (Eds.), Encyclopedia of machine learning (pp. 829–838). Springer. https://doi.org/10.1007/978-0-387-30164-8_705
[6]      Lee, C.-S., Wang, M.-H., & Hagras, H. (2010). A type-2 fuzzy ontology and its application to personal diabetic-diet recommendation. IEEE Transactions on Fuzzy Systems, 18(2), 374–395.
[7]      Ramachandran, S., Singh, H., & Bajaj, K. K. (1992). Nutrition diet programme – An expert system. In AAAI Fall Symposium Series (Technical Report FS-92-01). AAAI Press.
[8]      Chen, Y., Hsu, C.-Y., Liu, L., & Yang, S. (2012). Constructing a nutrition diagnosis expert system. Expert Systems with Applications, 39(2), 2132–2156.
[9]      Krbez, J. M., & Shaout, A. (2013). Fuzzy nutrition system. International Journal of Innovative Research in Computer and Communication Engineering, 1(7), 1–12.
[10]   Heinonen, P., Mannelin, M., Iskala, H., & Juuso, E. (2009). Development of a fuzzy expert system for a nutritional guidance application. In Proceedings of the Joint 2009 International Fuzzy Systems Association World Congress and 2009 European Society of Fuzzy Logic and Technology Conference (IFSA-EUSFLAT 2009). Lisbon, Portugal.
[11]   Iwendi, C., Khan, S., Anajemba, J. H., Bashir, A. K., & Noor, F. (2020). Realizing an efficient IoMT-assisted patient diet recommendation system through machine learning model. IEEE Access, 8, 28462–28474.
[12]   Harris, J. A., & Benedict, F. G. (1918). A biometric study of human basal metabolism. Proceedings of the National Academy of Sciences, 4(12), 370–373. https://doi.org/10.1073/pnas.4.12.370
[13]   Roza, A. M., & Shizgal, H. M. (1984). The Harris Benedict equation reevaluated: Resting energy requirements and the body cell mass. American Journal of Clinical Nutrition, 40(1), 168–182. https://doi.org/10.1093/ajcn/40.1.168
[14]   Mifflin, M. D., St Jeor, S. T., Hill, L. A., Scott, B. J., Daugherty, S. A., & Koh, Y. O. (1990). A new predictive equation for resting energy expenditure in healthy individuals. American Journal of Clinical Nutrition, 51(2), 241–247. https://doi.org/10.1093/ajcn/51.2.241
[15]   Frankenfield, D., Roth-Yousey, L., & Compher, C. (2005). Comparison of predictive equations for resting metabolic rate in healthy nonobese and obese adults: A systematic review. Journal of the American Dietetic Association, 105(5), 775–789. https://doi.org/10.1016/j.jada.2005.02.005
[16]   World Health Organization. (2000). Obesity: Preventing and managing the global epidemic (WHO Technical Report Series No. 894). World Health Organization.
[17]   Centers for Disease Control and Prevention, National Center for Health Statistics. (2009–2012). National Health and Nutrition Examination Survey data, 2009–2010 and 2011–2012 [Data set]. U.S. Department of Health and Human Services.
[18]   Pruim, R. (2015). NHANES: Data from the US National Health and Nutrition Examination Study (Version 2.1.0) [R package]. Project MOSAIC. https://github.com/ProjectMOSAIC/NHANES
[19]   FAO/WHO/UNU. (2004). Human energy requirements: Report of a joint FAO/WHO/UNU expert consultation (FAO Food and Nutrition Technical Report Series No. 1). Food and Agriculture Organization of the United Nations.
[20]   Trattner, C., & Elsweiler, D. (2017). Food recommender systems: Important contributions, challenges and future research directions (arXiv:1711.02760). arXiv. https://doi.org/10.48550/arXiv.1711.02760
[21]   Tran, T. N. T., Atas, M., Felfernig, A., & Stettinger, M. (2018). An overview of recommender systems in the healthy food domain. Journal of Intelligent Information Systems, 50(3), 501–526. https://doi.org/10.1007/s10844-017-0469-0
[22]   Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830.
[23]   American Diabetes Association. (2024). Standards of care in diabetes—2024. Diabetes Care, 47(Suppl. 1).
[24]   National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. (2002). Third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation, 106(25), 3143–3421. https://doi.org/10.1161/circ.106.25.3143
[25]   Jensen, M. D., Ryan, D. H., Apovian, C. M., Ard, J. D., Comuzzie, A. G., Donato, K. A., Hu, F. B., Hubbard, V. S., Jakicic, J. M., Kushner, R. F., Loria, C. M., Millen, B. E., Nonas, C. A., Pi-Sunyer, F. X., Stevens, J., Stevens, V. J., Wadden, T. A., Wolfe, B. M., & Yanovski, S. Z. (2014). 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults. Circulation, 129(25, Suppl. 2), S102–S138. https://doi.org/10.1161/01.cir.0000437739.71477.ee
[26]   Connor Gorber, S., Tremblay, M., Moher, D., & Gorber, B. (2007). A comparison of direct vs. self-report measures for assessing height, weight and body mass index: A systematic review. Obesity Reviews, 8(4), 307–326. https://doi.org/10.1111/j.1467-789X.2007.00347.x
[27]   National Research Council. (1989). Diet and health: Implications for reducing chronic disease risk. National Academy Press.
[28]   Kohavi, R. (1995). A study of cross-validation and bootstrap for accuracy estimation and model selection. In Proceedings of the 14th International Joint Conference on Artificial Intelligence (IJCAI) (Vol. 2, pp. 1137–1143). Morgan Kaufmann.
[29]   Cohen, J. (1968). Weighted kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychological Bulletin, 70(4), 213–220. https://doi.org/10.1037/h0026256
[30]   Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1, 206–215. https://doi.org/10.1038/s42256-019-0048-x
Transactions on Data Analysis in Social Science
Volume 8, Issue 2
Spring 2026
Pages 149-166

  • Receive Date 09 February 2026
  • Revise Date 25 April 2026
  • Accept Date 10 May 2026