AI-Driven Smart Food Ordering System with Personalized Nutrition Recommendations using Conversational Interface

Abstract

With the increasing popularity of online food ordering platforms, there remains a significant gap in delivering personalized and health-conscious food recommendations. This paper presents a Smart Food Ordering System that integrates AI-driven personalization based on individual user health data. The proposed system combines natural language processing (NLP) with FastText embeddings for intent classification and chatbot interaction, enabling users to place food orders through a conversational interface. Personalized recommendations are generated by analyzing user-specific health parameters such as age, weight, dietary restrictions, and fitness goals. The system is developed using Flask for the web interface and MongoDB for data storage, with additional modules for real-time order tracking, payment processing, and geolocation-based delivery validation. This integrated approach not only enhances user experience but also promotes healthier food choices. Experimental results demonstrate the system’s effectiveness in accurately understanding user intent and generating contextually relevant, health-optimized food recommendations.

Introduction

The rise of digital food ordering has transformed urban lifestyles by providing convenience and variety. However, current platforms often lack personalized recommendations that consider individual health needs like dietary restrictions and medical conditions. To address this, the paper proposes a Smart Food Ordering System that integrates AI and user health data to offer personalized, health-conscious meal suggestions via an interactive chatbot.

The system uses natural language processing with FastText embeddings to understand user intent and analyzes health data such as age, BMI, and dietary restrictions to tailor recommendations. Built with Flask and MongoDB, it includes modules for intent classification, AI-driven personalization, real-time order tracking, and geolocation-based delivery validation. A chatbot interface facilitates smooth, intuitive user interactions.

Performance evaluations show high intent classification accuracy (92.7%), fast chatbot response time (0.8 seconds), and relevant health-based recommendations (8.6/10 relevance score). The system also demonstrates low latency (1.2 seconds) and can support multiple concurrent users effectively. Compared to traditional platforms, this solution provides enhanced personalization, faster interaction, and promotes healthier food choices by integrating AI with user-centric health data.

Conclusion

This paper presented a Smart Food Ordering System that integrates AI-driven personalization based on user health data, aiming to bridge the gap between convenience and nutritional relevance in modern food delivery platforms. By combining FastText-based intent classification, an interactive chatbot interface, and health-aware food recommendation algorithms, the system offers a user-centric solution that promotes healthier eating habits. The integration of modules such as real-time order tracking, payment processing, and geolocation validation further enhances its usability and practicality in real-world scenarios. Experimental results demonstrated the system\'s effectiveness in terms of accuracy, response time, and recommendation relevance. The chatbot maintained a high response success rate with minimal latency, while the recommendation engine aligned closely with users’ dietary needs. These outcomes confirm the system’s potential for deployment in health-conscious and user-personalized food delivery ecosystems. While the current implementation offers promising results, several enhancements can be made to further improve functionality and user experience: 1) Integration with Wearable Devices: Real-time health metrics from fitness trackers and smartwatches can enable dynamic food recommendations based on current physical activity or medical readings. 2) Advanced Nutritional Models: Incorporating deep learning models and external nutrition databases could allow more precise and medically informed food suggestions. 3) Multilingual Support: Expanding NLP capabilities to support multiple languages will make the system accessible to a broader user base. 4) Scalability and Deployment: Hosting the system on a cloud infrastructure will improve scalability, allowing for larger-scale adoption across regions and user groups. 5) Feedback-Driven Learning: Implementing a feedback loop from users to refine the recommendation engine over time can help improve personalization accuracy and satisfaction.

References

[1] Hussain Alshahrani, Hanan Abdullah Mengash, Mashael Maashi, Fadoua Kouki, Ahmed Mahmud and Mesfer Al Duhayyim, ‘‘Enhancing Online Food Service User Experience Through Advanced Analytics and Hybrid Deep Learning for Comprehensive Evaluation’’ IEEE Access, vol. 12,pp [2] Jangmin Oh, ‘‘Developing a Model for Extracting Actual Product Names from Order Item Descriptions Using Generative Language Models,’’IEEE access, vol. 12,pp. 122695 - 122701, 2024 [3] Guido Rocchietti , Cosimo Rulli, Franco Maria Nardini, Cristina Ioana Muntean, Raffaele Perego, and Ophir Frieder, ‘‘ChatGPT Versus Modest Large Language Models: An Extensive Study on Benefits and Drawbacks for Conversational Search’’ieee access, vol. 12,pp. 15253 - 15271, 2025. [4] Raciel Yera Toledo, Ahmad A. Alzharani, AND Luis Martin, ‘A Food Recommender System Considering Nutritional Information and User Preferences,’IEEE access, vol. 12,pp.96695 - 96711, 2024 [5] S. Abinaya , K. S. Ashwin , AND A.Sherley Alphonse, Enhanced Emotion-Aware Conversational Agent: Analyzing User Behavioral Status for Tailored Reponses in Chatbot Interactions,’ IEEE access, vol. 13,pp.19770 - 19787, 2024 [6] https://fasttext.cc/docs/en/crawl-vectors.html [7] https://github.com/facebookresearch/fastText [8] https://www.nltk.org/ [9] G. Agapito, M. Simeoni, B. Calabrese, I. Caré, T. Lamprinoudi,P . H. Guzzi, A. Pujia, G. Fuiano, and M. Cannataro, ‘‘DIETOS: A dietary recommender system for chronic diseases monitoring and management,’’Comput. Methods Programs Biomed., vol. 153, pp. 93–104, Jan. 2018. [10] 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. [11] G. Williams, M. Tushev, F. Ebrahimi, and A. Mahmoud, ‘‘Modeling user concerns in sharing economy: The case of food delivery apps,’’ Automated Softw. Eng., vol. 27, nos. 3–4, pp. 229–263, Dec. 2020.

Copyright

Copyright © 2025 Bhuvaneshwari B, Vyshnavi J, Shanmuga Prabhaa S, Abdul Habeel. 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.