Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{199142,
        author = {Rahul Koli and Gopal Khorwal},
        title = {AI-BASED PERSONALIZED SUPERFOOD RECIPE RECOMMENDATION SYSTEM FOR NUTRITIONAL DEFICIENCY MANAGEMENT},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {13171-13190},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=199142},
        abstract = {Nutritional deficiencies remain a serious and largely preventable health problem affecting billions of people worldwide. Conditions like iron-deficiency anemia, low vitamin D, poor omega-3 intake, and insufficient dietary fiber are well-known contributors to chronic diseases including heart disease, type 2 diabetes, and weakened immunity. Yet despite the availability of dietary guidelines and nutrition apps, most existing tools offer the same generic advice to everyone, with no real connection to a person’s actual health data, food culture, or lifestyle. This paper presents the APSRRS — an AI-based Personalized Superfood Recipe Recommendation System — designed to address these gaps through intelligent, health-aware dietary planning. The system combines three AI approaches: collaborative filtering (learning from users with similar profiles), content-based filtering (analysing nutrient composition), and knowledge graph inference (understanding how nutrients interact). It reads the user’s blood panel data, dietary history, restrictions, and regional preferences, then generates complete personalised recipes targeting their specific deficiencies. A Random Forest classifier identifies deficiency risk, while an LSTM network tracks dietary patterns over time to keep recommendations relevant as the user’s health evolves. Tested on 1,200 user profiles across five deficiency categories, the system achieved a recommendation accuracy of 91.4%, a precision score of 0.89, and an average user satisfaction rating of 4.6 out of 5.0 — a 17.3% improvement over conventional rule-based systems. The results show that this kind of AI-driven, clinically grounded approach can make personalised nutrition support genuinely practical and effective at scale.},
        keywords = {Personalized Nutrition, Superfood Recommendation, Nutritional Deficiency Management, Hybrid Recommendation System, Machine Learning in Healthcare, Knowledge Graph, Collaborative Filtering, LSTM, Random Forest, Dietary Decision Support},
        month = {April},
        }