基于AIGC的营养配餐推荐系统研究
Research on Nutritional Meal Recommendation System Based on AIGC
陈钻凯 1王志林 1朱润键 1曾沛乐1
作者信息
- 1. 广州软件学院,广东 广州 510990
- 折叠
摘要
文章提出一种基于人工智能生成内容(AIGC)的营养配餐推荐系统,结合通义千问(Qwen)大语言模型和LORA技术对模型进行微调,以实现更精准的营养建议和配餐推荐.通过构建丰富的营养信息和菜谱数据的知识库,以及利用向量数据库进行高效检索,系统能够快速响应用户的查询请求,并提供个性化的配餐方案.实验证明,该系统在推荐准确性和用户体验上均优于传统方法.这项研究的贡献在于提出了一种新的营养配餐推荐方法,并通过实验证实了其有效性.
Abstract
This paper proposes a nutrition meal recommendation system based on Artificial Intelligence Generated Content(AIGC),which combines the Qwen large language model and LORA technology for fine-tuning to achieve more accurate nutrition advice and meal recommendations.By constructing a rich knowledge base of nutrition information and recipe data,and utilizing vector databases for efficient retrieval,the system can quickly respond to user queries and provide personalized meal plans.Experimental results demonstrate that the system outperforms traditional methods in recommendation accuracy and user experience.The contribution of this research lies in proposing a new method for nutrition meal recommendations and validating its effectiveness through experiments.
关键词
人工智能生成内容/营养配餐推荐系统/大语言模型/微调技术/向量数据库Key words
AIGC/nutrition meal recommendation system/large language model/fine-tuning technology/vector database引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据