基于矩阵分解和标签的饮食推荐系统

Diet Recommendation Based on Matrix Decomposition and Labels

食物品类的多样化导致选择更好和更健康的食物变得越来越复杂,目前大多数食物建议应用程序提供的只是一般的建议,而非根据用户的个人口味量身定制。为解决这一问题,本文提出一种饮食推荐系统,提供高质量和个性化的饮食建议。通过采集60位同学的饮食偏好数据生成饮食推荐结果,将用户偏好数据以用户评分和标签的形式表达,表明用户喜欢的食物成分或特征。实验结果表明,在饮食推荐算法中使用标签可以显著提高预测的准确性,预测的偏好能与用户真正的偏好相匹配。

The diversification of food categories has made choosing better and healthier foods increasingly complex,and most food suggestion apps today offer general advice rather than tailoring to the user’s personal tastes.To address this issue,this paper proposes a dietary recommendation system that provides high-quality and personalized dietary recommendations.The food recommendation results are generated by collecting the dietary preference data of 60 classmates,and the user preference data is expressed in the form of user scores and labels,indicating the food components or characteristics that users like.The experimental results show that the use of labels in the diet recommendation algorithm can significantly improve the accuracy of prediction,and the predicted preferences can match the real preferences of users.