计算机科学 ›› 2021, Vol. 48 ›› Issue (3): 113-118.doi: 10.11896/jsjkx.200900067

• 数据库&大数据&数据科学 • 上一篇    下一篇

面向协同过滤推荐的新型混合评分函数

肖诗涛1, 邵蓥侠1, 宋卫平2, 崔斌2   

  1. 1 北京邮电大学计算机学院 北京100876
    2 北京大学信息科学技术学院 北京100871
  • 收稿日期:2020-09-08 修回日期:2020-10-03 出版日期:2021-03-15 发布日期:2021-03-05
  • 通讯作者: 邵蓥侠(shaoyx@bupt.edu.cn)
  • 作者简介:stxiao@bupt.edu.cn
  • 基金资助:
    国家自然科学基金(U1936104,61702015);中央高校基本科研业务费专项资金资助(2020RC25)

Hybrid Score Function for Collaborative Filtering Recommendation

XIAO Shi-tao1, SHAO Ying-xia1, SONG Wei-ping2, CUI Bin2   

  1. 1 School of Computer Sicence,Beijing University of Posts and Telecommunications,Beijing 100876,China
    2 School of Electronics Engineering and Computer Science,Peking University,Beijing 100871,China
  • Received:2020-09-08 Revised:2020-10-03 Online:2021-03-15 Published:2021-03-05
  • About author:XIAO Shi-tao,born in 1998,postgra-duate,is a member of China Computer Federation.His main research interest is recommender system.
    SHAO Ying-xia,born in 1988,Ph.D,research associate professor,is a member of China Computer Federation.His main research interests include large-scale graph analysis,parallel computing framework and knowledge graph analysis.
  • Supported by:
    National Natural Science Foundation of China(U1936104,61702015) and Fundamental Research Funds for the Central Universities(2020RC25).

摘要: 协同过滤技术在现代推荐系统中得到了广泛的应用,其基本思想是相似的用户会喜欢相似的物品。评分函数(Score Function,SF)是协同过滤推荐模型的一个关键技术,用于评估用户对物品的喜好程度。然而,目前常用的评分函数存在如下缺陷,即内积评分函数难以有效捕捉用户与用户以及物品与物品的相似度,而欧几里德距离度量函数由于几何空间限制降低了模型的表达能力。文中提出了一种融合内积相似度和欧几里德距离度量的新颖的混合评分函数,并从理论上分析了此混合评分函数的性质,证明它能有效弥补现有评分函数的不足。此外,新的混合评分函数是一项通用技术,适用于诸多现有的推荐模型(如SVD++,MF,NGCF,CML等),能够提高模型的推荐质量。最后,在6个公开数据集上进行了大量实验,验证了新混合评分函数的优越性能。

关键词: 推荐系统, 协同过滤, 评分函数

Abstract: Collaborative Filtering has been widely used in modern recommendation systems,and it assumes that similar users prefer similar items.A key ingredient of CF-based recommendation model is the score function,which measures the preference of users on items.However,there are some shortages in the most popular score functions.The inner product score function fails to capture the user-user similarity and item-item similarity effectively,and Euclidean distance measurement function reduces the expressiveness of the model because of its geometrical restriction.This paper proposes a novel hybrid score function by mixing the inner product-based similarity and the Euclidean distance metric,and further theoretically analyze its properties,thus proving that the new score function can avoid the aforementioned shortages effectively.In addition,the new hybrid score function is a general technique and can help to improve the quality of recommendation for existing models (e.g.,SVD++,MF,NGCF,CML).Extensive empirical studies over 6 datasets demonstrate the superior performance of the proposed hybrid score function.

Key words: Recommendation system, Collaborative filtering, Score function

中图分类号: 

  • TP391.3
[1]YI X,YANG J,HONG L,et al.Sam-pling-bias-corrected neural modeling for large corpus item recommendations[C]//Procee-dings of ACM Conference on Recommender Systems.Copenha-gen:Association for Computing Machinery,2019:269-277.
[2]EBESU T,SHEN B,FANG Y.Collaborative Memory Network for Recommendation Systems[C]//Proceedings of the International ACM SIGIR Conference.Ann Arbor:Association for Computing Machinery,2018:515-524.
[3]KOREN Y,BELL R,VOLINSKY C.Matrix Factorization Tech-niques for Recommender Systems[J].Computer,2009,42(8):30-37.
[4]CHEN X,XU H,ZHANG Y,et al.Sequential Recommendation with User Memory Net-works[C]//Proceedings of the ACM International Conference on Web Search and Data Mining.Marina Del Rey:Association for Computing Machinery,2018:108-116.
[5]RENDLE S,FREUDENTHALER C,GANTNER Z,et al.BPR:Bayesian personalized ranking from implicit feedback[C]//Proceedings of the Conference on Uncertainty in Artificial Intelligence.Montreal:AUAI Press,2009:452-461.
[6]ZHENG L,LU C,JIANG F,et al.Spectral col-laborative filtering[C]//Proceedings of ACM Conference on Recommender Systems.Vancouver:Association for Computing Machinery,2018:311-319.
[7]HE X,LIAO L,ZHANG H,et al.Neural Collabora-tive Filtering[C]//Proceedings of the International Conference on World Wide Web.Perth:International World Wide Web Conferences Steering Committee,2017:173-182.
[8]HSIEH C,YANG L,CUI Y,et al.Collaborative Metric Lear-ning[C]//Proceedings of the International Conference on World Wide Web.Perth:International World Wide Web Conferences Steering Committee,2017:193-201.
[9]XIAO H,HUANG M,ZHU X.From one point to a manifold:knowledge graph embedding for precise link prediction[C]//Proceedings of the Conference on Uncertainty in Artificial Intelligence.New York:AUAI Press,2016:1315-1321.
[10]WANG X,HE X,WANG M,et al.Neural Graph Collaborative Filtering[C]//Proceedings of the International ACM SIGIR Conference.Paris:Association for Computing Machinery,2019:165-174.
[11]WANG H,WANG N,YEUNG D.Collaborative Deep Learning for Recommender Systems [C]//Proceedings of the ACM SIGKDD International Conference.Sydney:Association for Computing Machinery,2015:1235-1244.
[12]KHOSHNESHIN M,STREET W.Collaborative filtering viaeuclidean embedding[C]//Proceedings of ACM Conference on Recommender Systems.Barcelona:Association for Computing Machinery,2010:87-94.
[13]CEN Y,ZOU X,ZHANG J,et al.Representation Learning for Attributed Multiplex Heterogeneous Net-work[C]//Procee-dings of the ACM SIGKDD International Conference.Ancho-rage:Association for Computing Machinery,2019:1358-1368.
[14]MASSA P,AVESANI P.Trust-aware recom-mender systems[C]//Proceedings of ACM Conference on Recommender Systems.Minneapolis:Association for Computing Machinery,2007:17-24.
[15]ZIEGLER C,MCNEE S,KONSTAN J,et al.Improving recommendation lists through topic diversi-fication[C]//Proceedings of the International Conference on World Wide Web.Chiba:International World Wide Web Conferences Steering Committee,2005:22-32.
[16]MCAULEY J,TARGETT C,SHI Q,et al.Image-Based Recommendations on Styles and Substitutes[C]//Proceedings of the International ACM SIGIR Conference.Santiago:Association for Computing Machinery,2015:43-52.
[17]HE X,CHEN T,KAN M,et al.TriRank:Review-aware Ex-plainable Recommendation by Model-ing Aspects[C]//Procee-dings of the ACM International on Conference on Information and Knowledge Management.Melbourne:Association for Computing Machinery,2015:1661-1670.
[18]HERLOCKER J,KONSTAN J,BORCHERS A,et al.An algorithmic framework for performing collaborative filtering[C]//Proceedings of the International ACM SIGIR Conference.Berkeley:Association for Computing Machinery,1999:230-237.
[19]SARWAR B,KARYPIS G,KONSTAN J,et al.Item-based collaborative filtering recommendation algorithms[C]//Procee-dings of the International Conference on World Wide Web.Hong Kong:International World Wide Web Conferences Steering Committee,2001:285-295.
[20]KOREN Y.Factorization meets the neighborhood:a multiface-ted collaborative filtering model[C]//Proceedings of the ACM SIGKDD International Conference.Las Vegas:Association for Computing Machinery,2008:426-434.
[1] 郝志峰, 廖祥财, 温雯, 蔡瑞初. 基于多上下文信息的协同过滤推荐算法[J]. 计算机科学, 2021, 48(3): 168-173.
[2] 韩立锋, 陈莉. 融合用户属性与项目流行度的用户冷启动推荐模型[J]. 计算机科学, 2021, 48(2): 114-120.
[3] 马理博, 秦小麟. 话题-位置-类别感知的兴趣点推荐[J]. 计算机科学, 2020, 47(9): 81-87.
[4] 骆佳磊, 孟利民. 基于路口相似度的信号配时方案推荐算法[J]. 计算机科学, 2020, 47(6A): 66-69.
[5] 邹海涛, 郑尚, 王琦, 于化龙, 高尚. 基于牛顿法的自适应高阶评分距离推荐模型研究[J]. 计算机科学, 2020, 47(6A): 494-499.
[6] 朱磊, 胡沁涵, 赵雷, 杨季文. 基于评分偏好和项目属性的协同过滤算法[J]. 计算机科学, 2020, 47(4): 67-73.
[7] 赵楠, 皮文超, 许长桥. 一种面向多维特征分析过滤的视频推荐算法[J]. 计算机科学, 2020, 47(4): 103-107.
[8] 李太松,贺泽宇,王冰,颜永红,唐向红. 基于循环时间卷积网络的序列流推荐算法[J]. 计算机科学, 2020, 47(3): 103-109.
[9] 冯晨娇,梁吉业,宋鹏,王智强. 基于极端评分行为的相似度计算[J]. 计算机科学, 2020, 47(2): 31-36.
[10] 吴磊,岳峰,王含茹,王刚. 一种融合科研人员标签的学术论文推荐方法[J]. 计算机科学, 2020, 47(2): 51-57.
[11] 黄超然. 基于显式反馈协同过滤算法的偏好与共性平衡[J]. 计算机科学, 2020, 47(11A): 471-473.
[12] 罗鹏宇, 吴乐, 吕扬, 袁堃平, 洪日昌. 基于时序推理的分层会话感知推荐模型[J]. 计算机科学, 2020, 47(11): 73-79.
[13] 康雁, 卜荣景, 李浩, 杨兵, 张亚钏, 陈铁. 基于增强注意力机制的神经协同过滤[J]. 计算机科学, 2020, 47(10): 114-120.
[14] 王涵, 夏鸿斌. LDA模型和列表排序混合的协同过滤推荐算法[J]. 计算机科学, 2019, 46(9): 216-222.
[15] 邓存彬, 虞慧群, 范贵生. 融合动态协同过滤和深度学习的推荐算法[J]. 计算机科学, 2019, 46(8): 28-34.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 宫法明,李翛然. 基于Neo4j的海量石油领域本体数据存储研究[J]. 计算机科学, 2018, 45(6A): 549 -554 .
[2] 王顺,康达周,江东宇. 本体映射综述[J]. 计算机科学, 2017, 44(9): 1 -10 .
[3] 刘耀,帅远华,龚幸伟,黄毅. 基于领域本体的文本分割方法研究[J]. 计算机科学, 2018, 45(1): 128 -132 .
[4] 吴正洋,汤庸,方家轩,董浩业. 一种基于本体语义相似度的协同过滤推荐方法[J]. 计算机科学, 2015, 42(9): 204 -207 .
[5] 钟林辉,宗洪雁. 基于本体的构件化软件演化信息获取及度量研究[J]. 计算机科学, 2015, 42(1): 196 -200 .
[6] 郑志蕴,阮春阳,李伦,李钝. 本体语义相似度自适应综合加权算法研究[J]. 计算机科学, 2016, 43(10): 242 -247 .
[7] 柯昌博,黄志球,肖甫. 基于本体概念相似度的软件构件检索方法[J]. 计算机科学, 2017, 44(12): 144 -149 .
[8] 曾健荣, 张仰森, 郑佳, 黄改娟, 陈若愚. 面向多数据源的网络爬虫实现技术及应用[J]. 计算机科学, 2019, 46(5): 304 -309 .
[9] 许飞翔,叶霞,李琳琳,曹军博,王馨. 基于SA-BP算法的本体概念语义相似度综合计算[J]. 计算机科学, 2020, 47(1): 199 -204 .
[10] 蹇松雷, 卢凯. 复杂异构数据的表征学习综述[J]. 计算机科学, 2020, 47(2): 1 -9 .