AI Powered Health Nutrient Rating System

Abstract

Poor dietary habits contribute to numerous health issues globally each year. This often occurs due to limited nutritional awareness and the absence of accessible tools that help individuals evaluate the quality of their daily food intake. Current nutrition-monitoring practices typically treat nutrient assessment and dietary planning as separate processes, which can lead to inconsistent decisions and delayed lifestyle improvements. This document presents an integrated approach that combines automated nutrient analysis with intelligent meal-rating services using digital technology. By employing machine learning algorithms, the system examines factors such as calorie content, macronutrient distribution, micronutrient density, sugar levels, sodium intake, and harmful additives to determine the overall nutritional value of food items. When detecting unhealthy patterns, the software identifies suitable dietary adjustments and recommends expert-approved alternatives instantly. Merging predictive analytics with smart recommendation techniques enhances dietary management, supports timely health interventions, and improves user outcomes. Evidence indicates increased rating accuracy and more reliable nutritional guidance through AI-assisted evaluation systems. This demonstrates significant benefits in utilizing artificial intelligence for improving dietary efficiency and public health.

Introduction

This study presents an AI-Powered Health Nutrient Rating System designed to address the growing global burden of diet-related diseases such as obesity, metabolic syndrome, hypertension, and micronutrient deficiencies. Poor dietary habits remain a leading contributor to preventable morbidity and mortality worldwide. While existing nutrition-management systems provide dietary tracking or predictive analytics, they often operate in fragmented phases and lack real-time, automated intervention mechanisms. This disconnect between prediction and actionable guidance delays meaningful lifestyle corrections.

The proposed system bridges this gap by integrating machine learning–based nutrient evaluation with automated, real-time dietary recommendations. At its core, the framework employs an XGBoost-based predictive model to evaluate nutritional quality using variables such as calorie density, macronutrient balance, sodium, sugar, fiber, vitamin, and mineral content. When unhealthy patterns or high-risk nutrient scores exceed predefined thresholds, the system instantly generates personalized dietary recommendations and healthier alternatives without requiring manual expert consultation.

The architecture consists of a unified client–server ecosystem incorporating a React.js frontend, a Spring Boot backend, a MySQL database, and a machine learning engine. The system operates through four stages: input collection, nutrient prediction, decision analysis, and automated corrective action. A notification module delivers personalized alerts via email or SMS, enabling proactive intervention. Additionally, an appointment-scheduling module allows users to consult healthcare professionals when necessary, linking predictive insights with actionable healthcare access.

The literature review highlights two major but isolated advancements: (1) machine learning models capable of accurately predicting nutritional risk and food quality, and (2) digital diet-management platforms that enhance accessibility and adherence. However, most systems fail to merge these capabilities into a unified, risk-prioritized, automated nutrition-management framework. The proposed solution addresses this research gap by transforming predictive analytics into an intelligent, interactive health-guidance system.

Overall, the system shifts nutrition management from a reactive, analysis-only model to a preventive, real-time decision-support ecosystem, enhancing early dietary correction, personalization, scalability, and long-term health outcomes.

Conclusion

A new smart nutrient evaluation tool combined with automated food-rating workflows enhances responsiveness in user-centric dietary assessment systems. Integrating an advanced machine-learning algorithm into real-time nutritional scoring enhances the ability of the system to bridge the gap between analysing food quality and initiating prompt dietary recommendations. Empirical findings demonstrate this method’s efficacy through superior prediction precision, minimal processing delays, and robust user-friendliness; these metrics significantly enhance practical applications. An XGBoost-powered scoring system managed an accuracy rate of ninety-four percent. A precision of three percent and an error rate at zero point. A score of 0.96 in terms of Receiver Operating Characteristic Area Under Curve measurement. Its performance surpassed those of conventional machine-learning algorithms, thereby validating the efficacy of gradient boosting in evaluating nutritional health risks. The system’s automated recommendation module quickly arranges immediate guidance for individuals consuming unhealthy food items. It shortens the interval from detecting poor nutrition until receiving actionable suggestions. The method for forecasting dietary outcomes followed by instant guidance distinguishes this system from previous ML-based nutrition tools, many of which end prematurely in estimating nutrient values but fail to facilitate timely corrective interventions. Utilizing Java Spring Boot for development enhances capabilities in terms of system scalability, efficiency, and durability. In approximately three units of time, there is no significant delay in execution. In two seconds, the system enables seamless execution of urgent nutritional evaluation procedures. Introducing an all-purpose rating feature broadens its applicability in diverse dietary scenarios, excluding solely processed-food assessments. Nevertheless, there are certain constraints present within it. The success depends greatly upon having various types of food information and high-quality nutritional datasets readily accessible. Engaging in training on more extensive, practical food datasets will improve the model’s reliability and minimize errors. Although false positives pose lesser harm compared to false negatives when predicting dietary concerns, these may result in unnecessary user alerts. Therefore, it is imperative to employ adaptable thresholds alongside enhanced nutrient-risk assessment methodologies. Safety and confidentiality continue being vital because handling personal dietary information demands robust security measures such as encryption of communications and reliable user verification techniques. Potential enhancements could include integrating barcode scanners, Internet of Things-based smart kitchen gadgets, or ongoing sensor networks for immediate nutrient estimation. Enhanced neural networks integrating diverse datasets and decentralized machine learning techniques can enhance precision without compromising confidentiality. Extra components might broaden the scope of the system beyond its current emphasis on nutrition scoring to include conditions like obesity risks, metabolic disorders, gut-health monitoring, or persistent lifestyle-related issues. Combining intelligent food-triaging systems, remote dietician consultations via telehealth services, automatic trigger mechanisms for meal-planning alerts, and immediate notification systems can significantly improve nutritional decision-making. To summarize, this system seamlessly integrates machine-learning predictions into automated dietary guidance procedures, offering an efficient, health preserving approach in contemporary nutrition science. As enhancements continue, this system might evolve into an extensive intelligence-driven dietary monitoring tool capable of detecting nutritional imbalances at their earliest stages and offering proactive prevention strategies across vast populations.

References

[1] M. Chen, Y. Chen, … L. H.-IEEE T. M., and undefined 2019, “Smart food recognition using deep learning for automated diet monitoring,” ieeexplore.ieee.orgM Chen, Y Chen, L HoIEEE Transactions on Multimedia, 2019•ieeexplore.ieee.org, Accessed: Nov. 30, 2025. [Online].Available: https://ieeexplore.ieee.org/document/8700193/ [2] A. Salvador, N. Hynes, … T. M.-… C. V. P., and undefined 2017, “Learning cross-modal embeddings for cooking recipes and food images,” aclanthology.orgA Salvador, N Hynes, T MensinkCVPR. [3] S. Bharati, “Deep learning-based nutrient estimation from food images,” International Journal of Innovative Research in Computer Science and Technology (IJIRCST). [4] H. Jiang et al., “A survey of deep learning-based food image analysis: Recognition, portion estimation, and nutrition prediction,” mdpi.comH Jiang, L Li, 2023 [5] Bhagat, A. Vakil, R. Gupta, A. K.-P. C. Science, and undefined 2024, “Food classification using convolutional neural networks (CNN),” ElsevierD Bhagat, A Vakil, RK GuptaProcedia Computer Science, 2024 [6] N. Yalçin, M. A.- Heliyon, and undefined 2024, “Introducing a novel dataset for fine grained food recognition and demonstrating performance gains via preprocessing,” [7] S. Zhang, Y. Yang, C. Chen, X. Zhang, … Q. L.-E. S. with, and undefined 2024, “Deep learning-based multimodal dietary analysis from images, ingredients, and text: A systematic review,” ElsevierS Zhang, Y YangExpert Systems with Applications, 2024 [8] S. Taware, A. T.-Circuits, Systems, and Signal Processing, and undefined 2025, “Multimodal food recognition combining CNN and LSTM for nutrient estimation,” SpringerS Taware, AD ThakareCSSP, 2025 [9] N. Reimers and I. Gurevych, “Sentence-BERT: Sentence embeddings for ingredient and recipe text understanding,” EMNLP-IJCNLP 2019, pp. 3982–3992, 2019, doi: 10.18653/v1/D19-1410. [10] Z. W.-D. L.-B. M. S. Analysis and undefined 2024, “Deep learning-based multimodal food sentiment and healthiness analysis,” scitepress.orgZ WangDL-Based Multimodal Nutrition Analysis, 2024•scitepress.org, 2024 [11] “(Pan et al., 2023 [Multimodal Food Assessment using Image and Ingredients] - Google Scholar.” Accessed: Nov. 30, 2025 [12] S. Zhang, Y. Yang, C. Chen, X. Zhang, … Q. L.-E. S. with, and undefined 2024, “Deep learning-based multimodal nutrition recognition from images, metadata, and recipe text,” ElsevierExpert Systems with Applications, 2024 [13] “(Islam et al., 2023 [Multimodal Food and Nutrition Analysis Using Deep Learning] - Google Scholar.” Accessed: Nov. 30, 2025 [14] S. R. Jagtap, V. Kulkarni, Y. Pachorkar, O. Taur, and S. Gupta, “AI-driven real-time dietary evaluation app using food image analysis,” [15] S. Bharati, “Deep learning-based nutrient estimation using AI,” IJIRCST, no. 13, pp. 2347–5552, 2025, doi: 10.55524/ijircst.2025.13.5.6. [16] P. J. Uppalapati, M. Dabbiru, and V. R. Kasukurthi, “Investigating dietary habits using a novel dataset of university students,” Springer, vol. 5, no. 8, Aug. 2025, doi: 10.1007/S43545-025-01143-4. [17] A. Salvador, N. Hynes, S. A.-ACL, and undefined 2020, “Cross-modal food analytics for automated nutrition estimation,” ACL, pp. 46–54, 2020, doi: 10.18653/v1/2020.challengehml-1.6. [18] D. Bhagat, A. Vakil, R. K. Gupta, and A. Kumar, “Food image categorization using CNN,” Procedia Comput Sci, vol. 235, pp. 2079–2089, Jan. 2024, doi: 10.1016/J.PROCS.2024.04.197. [19] N. Yalçin and M. Alisawi, “A novel dataset for food recognition and enhanced preprocessing techniques,” Heliyon, vol. 10, no. 20, p. e38913, Oct. 2024, doi: 10.1016/J.HELIYON.2024.E38913. [20] S. Taware, A. D. Thakare, S. Taware, and A. D. Thakare, “Multimodal food recognition combining visual and textual cues,” CSSP, vol. 44, no. 9, pp. 6622–6649, Sep. 2025, doi: 10.1007/S00034-025-03080-2. [21] N. Reimers and I. Gurevych, “Sentence-BERT for nutritional text and ingredient analysis,” EMNLP-IJCNLP, 2019, doi: 10.18653/V1/D19-1410. [22] Z. Wang, “Deep Learning-Based Multimodal Food Analysis,” 2024, doi: 10.5220/0012842200004547. [23] H. Jiang, C. Lu, S. Li, Y. Zhao, C. Tang, and Y. Zong, “A survey of deep learning-based multimodal dietary assessment,” Entropy, vol. 25, no. 10, p. 1203, Oct. 2023, doi: 10.3390/E25101203. [24] J. Pan, W. Fang, Z. Zhang, … B. C.-I. O. J. of, and undefined 2023, “Multimodal nutrient recognition using images, OCR text, and metadata,” IEEE Open Journal of Engineering in Medicine and Biology, 2023•ieeexplore.ieee.org, Accessed: Nov. 30, 2025. [Online]. [25] H. Jiang et al., “A survey of deep learning-based food and nutrient analysis,” Entropy, vol. 25, no. 10, p. 1203, Oct. 2023, doi: 10.3390/E25101203. [26] J. Pan, W. Fang, Z. Zhang, B. Chen, Z. Zhang, and S. Wang, “Multimodal food analysis combining visual and OCR data,” IEEE Open J Eng Med Biol, vol. 5, p. 396, 2023, doi: 10.1109/OJEMB.2023.3240280. [27] A. Salvador, N. Hynes, S. A., and undefined 2020, “Multimodal diet analytics and automated nutrient scoring,” ACL, pp. 46–54, 2020, doi: 10.18653/V1/2020.CHALLENGEHML-1.6. [28] M. Bin Habib, M. Ferdous, N. A. Khan, and S. Hossain, “Multimodal nutrition analysis using fusion techniques and transformers,” IJACSA, vol. 15, no. 6, 2024, Accessed: Dec. 01, 2025. [Online]. Available: www.ijacsa.thesai.org. [29] “(PDF) Multimodal Analysis For Nutrient Assessment.” Accessed: Dec. 01, 2025. [Online]. Available: https://www.researchgate.net/publication/ [30] N. Yalçin, M. A.-Heliyon, and undefined 2024, Introducing a dataset for food recognition with preprocessing enhancements,” cell.comN Yalçin, M AlisawiHeliyon, 2024•cell.com, Accessed: Dec. 01, 2025. [Online]. Available: https://www.cell.com/heliyon/fulltext/ [31] R. Mandal, P. Lohar, D. Patil, A. Patil, and S. Wagh, “AI-based dietary evaluator using image recognition and nutrient estimation,” ICISCoIS 2023, pp. 521–526, 2023, doi: 10.1109/ICISCOIS56541.2023.10100589. [32] J. M. Fernández-Dols and R. García-González, “Food perception and multimodal sensory evaluation,” Springer, pp. 399–424, Jan. 2025, doi: 10.1007/978-3-031-700644_16.

Copyright

Copyright © 2026 Harjot Singh, Gopal Yadav, Jatin Kumar, Priyanka Gupta . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.