Satish Singh Mekale; Maumita Chakraborty; Chiradeep C. Mukherjee

Explainable AI for Diabetes Nutrition: Time-Aware Seq2Seq Learning for Personalized 21-Meal Weekly Planning

PDF (1214KB), PP.151-170

Views: 0 Downloads: 0

Author(s)

Satish Singh Mekale ^1,* Maumita Chakraborty ¹ Chiradeep C. Mukherjee ¹

1. Institute of Engineering & Management Kolkata, University of Engineering and Management, Kolkata, 700091, India

* Corresponding author.

DOI: https://doi.org/10.5815/ijitcs.2026.05.10

Received: 11 Mar. 2026 / Revised: 7 May 2026 / Accepted: 23 Jul. 2026 / Published: 8 Oct. 2026

Index Terms

Type 2 Diabetes Mellitus, Personalized Meal Planning, Explainable Artificial Intelligence, Time-aware Recommendation Systems, Long Short Term Memory Networks, Nutritional Constraint Satisfaction, Health-Aware Recommender Systems

Abstract

Adequate nutrition, patient acceptance and consistent dietary adherence are required for optimal chronic glycaemic control in T2DM. Customised and clinically adapted meal planning may best support these factors. This work is situated in the field of Medical Informatics, combining explainable sequence modelling with clinically meaningful dietary restrictions, ensuring both prediction accuracy and health-oriented decision support for Type 2 Diabetes Mellitus (T2DM). In contrast to the traditional RNN based recommenders, the proposed approach explicitly incorporates adherence-aware assessment, interpretability and nutritional safety. In this paper, we propose CA-Seq2Seq-LSTM-Attn, a Constraint-Aware Seq2Seq-LSTM with Dot-Product Attention framework for generating customised seven-day meal plans (21 meals) from nutritional constraints and previous dietary behaviour. We use an LSTM-based Seq2Seq architecture in place of Transformer-based models because of its strong inductive bias for modelling short, structured temporal sequences (21 meals over 7 days), better generalisation on small high-variance dietary datasets, and lower computational complexity than data-hungry Transformer models. The model is a sequence-to-sequence Long Short-Term Memory (LSTM) architecture with attention for capturing temporal eating patterns in daily meal sequences. The constraint-aware decoding strategy ensures that the calorie, carbohydrate, sugar, protein limits, and the Glycaemic Index (GI) and Glycaemic Load (GL) thresholds are satisfied, which is crucial in the context of diabetes management. The attention mechanism makes the respective forecast more interpretable by pointing to important previous meals. We evaluate the system using enriched user-recipe interactions and comprehensive nutritional data from the Food.com dataset. The experimental results show a good ranking performance with NDCG@10 of 0.5026 and Recall@10 of 0.5909. The model also ensures a better diet consistency, as indicated by the lower variance in carbs (826.6690), weekday adherence (74.39%), intra-plan variety (0.7180), and high explainability (minimum perturbation-based faithfulness confidence reduction: −0.0288). Incorporating temporal sequence modelling, attention-based interpretability and constraint-aware decoding substantially improve clinical nutritional feasibility and recommendation accuracy. The proposed framework outperforms the traditional Collaborative Filtering (BPR-MF, NCF) and sequential deep learning baselines (GRU4Rec, SASRec) on all ranking and nutrition metrics. The proposed CA-Seq2Seq-LSTM-Attn framework is a promising decision-support tool for personalised dietary control in T2DM as it generates interpretable, nutritionally balanced and clinically matched meal recommendations.

Cite This Paper

Satish Singh Mekale, Maumita Chakraborty, Chiradeep C. Mukherjee, "Explainable AI for Diabetes Nutrition: Time-Aware Seq2Seq Learning for Personalized 21-Meal Weekly Planning", International Journal of Information Technology and Computer Science(IJITCS), Vol.18, No.5, pp.151-170, 2026. DOI:10.5815/ijitcs.2026.05.10

Reference

[1]I. M. De Hoogh, M. J. Reinders, L. Doets, F. P. M. Hoevenaars, and L. Jan, “Design Issues in Personalized Nutrition Advice Systems Corresponding Author :,” J. Med. INTERNET Res., vol. 25, 2023, doi: 10.2196/37667.
[2]S. K. Aydın, R. H. Ali, S. Faiz, and T. A. Khan, “An Integrated AI Framework for Personalized Nutrition Using Machine Learning and Natural Language Processing for Dietary Recommendations,” 2025. doi: 10.3390/app15179283.
[3]O. Chávez-Bosquez, J. Marchi, and P. Pozos, “Nutritional Menu Planning: A Hybrid Approach and Preliminary Tests,” Res. Comput. Sci., vol. 82, pp. 93–104, Nov. 2014, doi: 10.13053/rcs-82-1-8.
[4]R. Fox and Y. Bui, “An Artificial Intelligence Approach to Nutritional Meal Planning for Cancer Patients,” Adv. Intell. Syst. Comput., vol. 347, pp. 215–224, Jan. 2015, doi: 10.1007/978-3-319-18476-0_22.
[5]V. Espín, M. V Hurtado, and M. Noguera, “Nutrition for Elder Care: a nutritional semantic recommender system for the elderly,” Expert Syst., vol. 33, no. 2, pp. 201–210, Apr. 2016, doi: https://doi.org/10.1111/exsy.12143.
[6]T. Cioara et al., “Expert system for nutrition care process of older adults,” Futur. Gener. Comput. Syst., vol. 80, pp. 368–383, 2018, doi: https://doi.org/10.1016/j.future.2017.05.037.
[7]M. Amiri, F. Sarani Rad, and J. Li, “Delighting Palates with AI: Reinforcement Learning’s Triumph in Crafting Personalized Meal Plans with High User Acceptance,” 2024. doi: 10.3390/nu16030346.
[8]K. Kalpakoglou et al., “An AI-based nutrition recommendation system: technical validation with insights from Mediterranean cuisine.,” Front. Nutr., vol. 12, p. 1546107, 2025, doi: 10.3389/fnut.2025.1546107.
[9]D. Tsolakidis, L. P. Gymnopoulos, and K. Dimitropoulos, “Artificial Intelligence and Machine Learning Technologies for Personalized Nutrition: A Review,” 2024. doi: 10.3390/informatics11030062.
[10]C. Liu et al., “A New Deep Learning-Based Food Recognition System for Dietary Assessment on An Edge Computing Service Infrastructure,” IEEE Trans. Serv. Comput., vol. 11, no. 2, pp. 249–261, 2018, doi: 10.1109/TSC.2017.2662008.
[11]V. Sharma, V. Sharma, A. Khan, and D. J. Wassmer, “Malnutrition , Health and the Role of Machine Learning in Clinical Setting,” Front. Nutr., vol. 7, no. April, pp. 1–9, 2020, doi: 10.3389/fnut.2020.00044.
[12]A. Barredo Arrieta et al., “Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI,” Inf. Fusion, vol. 58, pp. 82–115, 2020, doi: https://doi.org/10.1016/j.inffus.2019.12.012.
[13]F. Di Martino, F. Delmastro, and C. Dolciotti, “Explainable AI for malnutrition risk prediction from m-Health and clinical data,” Smart Heal., vol. 30, p. 100429, 2023, doi: https://doi.org/10.1016/j.smhl.2023.100429.
[14]K. Chadaga, S. Prabhu, N. Sampathila, and R. Chadaga, “A machine learning and explainable artificial intelligence approach for predicting the efficacy of hematopoietic stem cell transplant in pediatric patients,” Healthc. Anal., vol. 3, p. 100170, 2023, doi: https://doi.org/10.1016/j.health.2023.100170.
[15]G. M. Setegn, “Improving machine learning models through explainable AI for predicting the level of dietary diversity among Ethiopian preschool children,” Ital. J. Pediatr., vol. 1, pp. 1–11, 2025, doi: https://doi.org/10.1186/s13052-025-01892-1.
[16]J. Zhang, Z. Wang, W. Liu, X. Liu, and Q. Zheng, “A unified approach to designing sequence-based personalized food recommendation systems: tackling dynamic user behaviors,” Int. J. Mach. Learn. Cybern., vol. 14, no. 9, pp. 2903–2912, 2023, doi: 10.1007/s13042-023-01808-7.
[17]M. S. N. Siemon, A. S. M. Shihavuddin, and G. Ravn-Haren, “Sequential transfer learning based on hierarchical clustering for improved performance in deep learning based food segmentation,” Sci. Rep., vol. 11, no. 1, p. 813, 2021, doi: 10.1038/s41598-020-79677-1.
[18]N. R. Tran, Y. Zhang, R. M. Leech, and S. A. McNaughton, “Predicting diet quality and food consumption at eating occasions using contextual factors: an application of machine learning models.,” Int. J. Behav. Nutr. Phys. Act., vol. 22, no. 1, p. 136, Nov. 2025, doi: 10.1186/s12966-025-01818-4.
[19]M. Rostami, M. Oussalah, and V. Farrahi, “A Novel Time-Aware Food Recommender-System Based on Deep Learning and Graph Clustering,” IEEE Access, vol. 10, pp. 52508–52524, 2022, doi: 10.1109/ACCESS.2022.3175317.
[20]I. Papastratis, D. Konstantinidis, P. Daras, and K. Dimitropoulos, “AI nutrition recommendation using a deep generative model and ChatGPT,” Sci. Rep., vol. 14, Jun. 2024, doi: 10.1038/s41598-024-65438-x.
[21]S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, 1997, doi: 10.1162/neco.1997.9.8.1735.
[22]S. Hochreiter and J. Schmidhuber, “LSTM can solve hard long time lag problems,” Adv. Neural Inf. Process. Syst., vol. 9, 1996.
[23]G. Van Houdt, C. Mosquera, and G. Nápoles, “A review on the long short-term memory model,” Artif. Intell. Rev., vol. 53, no. 8, pp. 5929–5955, 2020, doi: 10.1007/s10462-020-09838-1.
[24]Colah, “Understanding LSTM Networks,” 2015.
[25]X.-H. Le, H. V Ho, G. Lee, and S. Jung, “Application of Long Short-Term Memory (LSTM) Neural Network for Flood Forecasting,” 2019. doi: 10.3390/w11071387.
[26]M. Rostami, V. Farahi, K. Berahmand, S. Forouzandeh, S. Ahmadian, and M. Oussalah, “A Novel Explainable and Health-aware Food Recommender System,” Int. Jt. Conf. Knowl. Discov. Knowl. Eng. Knowl. Manag. IC3K - Proc., vol. 3, no. Ic3k, pp. 208–215, 2022, doi: 10.5220/0011561700003335.
[27]S. Forouzandeh, M. Rostami, K. Berahmand, and R. Sheikhpour, “Health-aware food recommendation system with dual attention in heterogeneous graphs,” Comput. Biol. Med., vol. 169, no. May 2023, p. 107882, 2024, doi: 10.1016/j.compbiomed.2023.107882.
[28]M. Rostami, V. Farrahi, S. Ahmadian, S. Mohammad Jafar Jalali, and M. Oussalah, “A novel healthy and time-aware food recommender system using attributed community detection,” Expert Syst. Appl., vol. 221, no. February, p. 119719, 2023, doi: 10.1016/j.eswa.2023.119719.

International Journal of Information Technology and Computer Science (IJITCS)