中北大学学报（自然科学版）2018，Vol.39Issue(5)：529-535,7.DOI:10.3969/j.issn.1673-3193.2018.05.008
用户-电影网中基于属性和用户偏好的推荐方法比较研究
Comparative Study on the Recommendation Methods Based on Attributes and User Preferences on the User-Movie Network
摘要
Abstract
The item attributes and user preferences were not considered in the classical algorithms,which were the item-based collaborative filtering and the user-based collaborative filtering.Based on these two points,two new recommendation methods,the item attribute-based recommendation and the user pref-erence-based recommendation were proposed to apply to the user-movie network.The methods proposed were compared with the classical recommendation methods according to the metrics:recommendation precision,recall,novelty and diversity.The simulation results show that the performances of the rec-ommendation methods are related to the length of the recommendation list.With the increase of the length of the recommendation list,the recommendation recall and novelty of the four methods are on the rise,while the recommendation diversity is on the decline;in terms of recommendation novelty and di-versity,the proposed methods are superior to the classical recommendation methods;as for precision, the recommendation precision of the other three methods is decreasing with the increase of the length of the recommendation list,while the precision of the user preference-based recommendation increases first and then decreases,and the effectiveness is better.On the whole,the two recommendation methods proposed can effectively and accurately recommend suitable items to users.关键词
推荐系统/协同过滤/用户-电影网/物品属性/用户偏好Key words
recommender system/collaborative filtering/user-movie network/item attributes/user preferences分类
信息技术与安全科学引用本文复制引用
张舒娟,靳祯..用户-电影网中基于属性和用户偏好的推荐方法比较研究[J].中北大学学报（自然科学版）,2018,39(5):529-535,7.基金项目
国家自然科学基金资助项目(11331009) （11331009）
