摘要
设计应用于自助点餐平台的推荐系统模型,将用户的协同过滤推荐应用到推荐系统中,根据获取用户的点菜频率和用户查看菜单的模式构建用户—菜品评价矩阵,在计算用户对菜品兴趣度时增加季节作为权重系数,提高菜品推荐准确性,在推荐过程中通过获取用户特征信息和给菜品打标签解决冷启动的问题,并结合关联规则实现套餐推荐.
Designing recommendation system model applied to self-service ordering platform adopting the user-based collaborative filtering.Generating user-dish evaluation matrix according to getting the user ordering frequency and the mode of the user viewing menu.When calculate the user＇s interest in the dish increased season as weight coefficient in order to improve the recommendation accuracy.Through getting user features information and labelling the dish can solve the cold start problem in the process of recommendation,and combining association rules is to achieve package recommendation.
出处
《漳州师范学院学报（自然科学版）》
2013年第2期32-35,共4页
Journal of ZhangZhou Teachers College（Natural Science)
基金
2010年国家自然科学基金项目(61070151)
2010厦门市科技计划指导性项目(3502Z20109003)
关键词
自助点餐
协同过滤
关联规则
推荐系统
self-service ordering
collaborative filtering
association rules
recommendation system
