一种基于粒计算的美食推荐方法  

A Catering Recommendation Method Based on Granular Computing

作　　者：黄桂吉 刘盾[1] 叶晓庆 HUANG Guiji;LIU Dun;YE Xiaoqing(School of Economics and Management,Southwest Jiaotong University,Chengdu 610031,China;School of Computing and Artificial Intelligence,Southwest Jiaotong University,Chengdu 611756,China)

机构地区：[1]西南交通大学经济管理学院,四川成都610031 [2]西南交通大学计算机与人工智能学院,四川成都611756

出　　处：《昆明理工大学学报（自然科学版）》2023年第3期93-104,共12页Journal of Kunming University of Science and Technology(Natural Science)

基　　金：国家自然科学基金项目(62276217,61876157,71571148);四川省杰出青年科学基金项目(2022JDJQ0034);重庆市计算智能重点实验室项目(2020FF03);中央高校基本科研基金项目(2682022CX063,2682022ZTPY057)。

摘　　要：推荐系统一直以来都是美食推荐关注的重要研究方向.然而,现有推荐方法大多数都以静态文本挖掘为主,缺乏对推荐动态变化的思考.为了体现序贯决策的美食筛选过程,本研究将粒计算思想引入到基于美食的电子商务推荐中.首先,针对美食推荐动态性特征,将用户的兴趣偏好通过粒化的方式进行描述和解释,形成不同层次的属性集合序列;其次,根据用户推荐需求,在各层次间自由选择相应的属性集合序列并计算分析,得到美食推荐中的属性重要度排序以及最优粒度和相应的推荐结果列表;最后,将大众点评美食推荐作为应用场景,对本文所提方法的有效性和可行性进行实验验证.The recommendation system has always been an important research direction for catering recommendation.However,most existing recommendation methods mainly focus on static text mining,lacking consideration for the dynamic changes in recommendations.To reflect the sequential decision-making process in food selection,this study introduces the concept of granular computing into food-based e-commerce recommendations.First,to address the dynamic characteristics of catering recommendations,users'interest preferences are described and explained through granularization,forming attribute sets sequences at different levels.Next,according to user recommendation needs,the corresponding attribute set sequences are freely selected at each level for calculation and analysis.This yields the ranking of attribute importance and the optimal granularity in catering recommendations,as well as the corresponding recommendation result lists.Finally,this study uses Dianping recommendations as an application scenario,to experimentally validate the effectiveness and feasibility of the proposed method.

关 键 词：粒计算 三支决策 多属性决策 属性权重 推荐系统 

分 类 号：TS971[轻工技术与工程] TP391.3[自动化与计算机技术—计算机应用技术]