ChatDiet: Empowering Personalized Nutrition-Oriented Food Recommender Chatbots through an LLM-Augmented Framework

Zhongqi Yang,Elahe Khatibi,Nitish Nagesh,Mahyar Abbasian,Iman Azimi,Ramesh Jain,Amir M. Rahmani

DOI: https://doi.org/10.1016/j.smhl.2024.100465

2024-09-25

Abstract:The profound impact of food on health necessitates advanced nutrition-oriented food recommendation services. Conventional methods often lack the crucial elements of personalization, explainability, and interactivity. While Large Language Models (LLMs) bring interpretability and explainability, their standalone use falls short of achieving true personalization. In this paper, we introduce ChatDiet, a novel LLM-powered framework designed specifically for personalized nutrition-oriented food recommendation chatbots. ChatDiet integrates personal and population models, complemented by an orchestrator, to seamlessly retrieve and process pertinent information. The personal model leverages causal discovery and inference techniques to assess personalized nutritional effects for a specific user, whereas the population model provides generalized information on food nutritional content. The orchestrator retrieves, synergizes and delivers the output of both models to the LLM, providing tailored food recommendations designed to support targeted health outcomes. The result is a dynamic delivery of personalized and explainable food recommendations, tailored to individual user preferences. Our evaluation of ChatDiet includes a compelling case study, where we establish a causal personal model to estimate individual nutrition effects. Our assessments, including a food recommendation test showcasing a 92\% effectiveness rate, coupled with illustrative dialogue examples, underscore ChatDiet's strengths in explainability, personalization, and interactivity.

Information Retrieval,Artificial Intelligence,Machine Learning,Multimedia

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to address several core issues in current nutrition-oriented food recommendation services:

1. **Personalization**:
- Current systems often lack true personalization capabilities, making it difficult to tailor food suggestions based on an individual's specific nutritional needs.
- Existing nutrition recommendations mainly rely on population-based standards, ignoring subtle differences in nutritional responses among individuals.

2. **Explainability**:
- Most existing systems rely on "black-box" machine learning models, which cannot clearly explain the logic behind their recommendations.
- It is difficult to provide detailed explanations tailored to an individual's specific nutritional needs.

3. **Interactivity**:
- Current food recommendation systems lack the ability to dynamically respond to user feedback, making it impossible to adjust recommendations in real-time to accommodate users' latest preferences or changes in status (such as temporary dietary restrictions).

To address these issues, the paper proposes a new framework called ChatDiet, which leverages large language models (LLMs) to enhance personalization, explainability, and interactivity. Specifically:

- **Personalized Model**: Evaluates the personalized nutritional effects for specific users through causal discovery and reasoning techniques.
- **Population Model**: Provides general information about the nutritional content of foods.
- **Coordinator**: Responsible for seamlessly retrieving and processing relevant information from both the personal model and the population model, and integrating it before passing it to the LLM to generate customized food recommendations.

In this way, ChatDiet can provide dynamic and personalized food recommendations, supporting users in achieving specific health goals. Additionally, the paper demonstrates the effectiveness of ChatDiet through case studies and illustrates its advantages in explainability, personalization, and interactivity with dialogue examples.