广东工业大学学报 ›› 2018, Vol. 35 ›› Issue (04): 45-50.doi: 10.12052/gdutxb.170140

• 综合研究 • 上一篇    下一篇

融合社交和标签信息的隐语义模型推荐算法

彭嘉恩1, 邓秀勤1, 刘太亨1, 刘富春2, 李文洲1   

  1. 1. 广东工业大学 应用数学学院, 广东 广州 510520;
    2. 广东工业大学 计算机学院, 广东 广州 510006
  • 收稿日期:2017-10-10 出版日期:2018-07-09 发布日期:2018-05-24
  • 通信作者: 邓秀勤(1966-),女,教授,主要研究方向为智能计算、数据挖掘等.E-mail:xiuqindeng@163.com E-mail:xiuqindeng@163.com
  • 作者简介:彭嘉恩(1993-),女,硕士研究生,主要研究方向为数据挖掘、机器学习、推荐系统.
  • 基金资助:
    国家自然科学基金资助项目（61673122，61273118）；广东省公益研究与能力建设专项资金资助项目（2015A030402006）；广东工业大学研究生创新创业及竞赛资助项目（2016YJSCX036，2017YJSCX039）

A Recommendation Algorithm of Latent Factor Model Fused with the Social and Tag Information

Peng Jia-en1, Deng Xiu-qin1, Liu Tai-heng1, Liu Fu-chun2, Li Wen-zhou1   

  1. 1. School of Applied Mathematics, Guangdong University of Technology, Guangzhou 510520, China;
    2. School of Computers, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2017-10-10 Online:2018-07-09 Published:2018-05-24

HTML

PDF (PC)

2801

摘要： 为了提高隐语义模型在数据稀疏情况下推荐结果的质量，提出一种带有社交正则化项和标签正则化项的隐语义模型.根据用户社交网络和物品标签的信息，设计出描述用户和物品概况的正则化项，并利用用户对物品的历史评分计算得到用户评分偏好，将这三项引入矩阵分解目标函数中，进一步约束目标函数，最后通过梯度下降法去优化模型参数，得到推荐结果.为了验证算法的有效性，在Last.fm数据集上进行实验，实验结果表明，本文算法的推荐质量优于其他传统推荐算法.

关键词: 隐语义模型, 社交网络, 标签信息, 推荐算法

Abstract: In order to improve the recommendation performance of latent factor model under the circumstance of data sparseness, a latent factor model with the social regularization and the tag regularization is proposed. According to the user's social network and the item's tag information, the regularization depicting the profiles of the user and the item is designed, and the user rating preferences calculated by using user's history rating of items. These three items are introduced into the objective function of the matrix decomposition to further constrain the objective function. Finally, the gradient descent method is used to optimize the model parameters and get the recommendation result. To verify the efficacy of the proposed method, the model is tested by the Last.fm data set, and the experimental results show that the recommendation algorithm proposed in this study has a better recommendation performance compared with other traditional recommendation algorithms.

Key words: latent factor model, social network, tag information, recommendation algorithm

中图分类号: 

  • TP312
[1] MA T, ZHOU J, TANG M, et al. Social network and tag sources based augmenting collaborative recommender system[J]. Ieice Transactions on Information & Systems, 2015, 98(4):902-910.
[2] SHAMBOUR Q, LU J. An effective recommender system by unifying user and item trust information for B2B applications[J]. Journal of Computer and System Science, 2015, 81(7):1110-1126.
[3] ZHENG N, LI Q. A recommender system based on tag and time information for social tagging systems[J]. Expert Systems with Applications, 2011, 38(4):4575-4587.
[4] MA H, ZHOU D, LIU C, et al. Recommender systems with social regularization[C]//Forth International Conference on Web Search and Web Data Mining. Hong Kong:ACM, 2011:287-296.
[5] LIU X, ABERER K. SoCo:A social network aided context-aware recommender system[C]//International Conference on World Wide Web. Rio de Janeiro:ACM, 2013:781-802.
[6] MA H, KING I, LYU M R. Learning to recommend with social trust ensemble[C]//International ACM SIGIR Conference on Research and Development in Information Retrieval. Boston:ACM, 2009:203-210.
[7] MA H, KING I, LYU M R. Learning to recommend with explicit and implicit social relations[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3):1-19.
[8] QIAN F, ZHAO S, TANG J, et al. SoRS:social recommendation using global rating reputation and local rating similarity[J]. Physica A Statistical Mechanics & Its Applications, 2016, 461(10):61-72.
[9] KIM K R, MOON N M. Recommender system design using movie genre similarity and preferred genres in Smart Phone[J]. Multimedia Tools and Applications, 2012, 61(1):87-104.
[10] CHOI S M, KO S K, HAN Y S. A movie recommendation algorithm based on genre correlations[J]. Expert Systems with Applications, 2012, 39(9):8079-8085.
[11] MA T, SUO X, ZHOU J, et al. Augmenting matrix factorization technique with the combination of tags and genres[J]. Physica A:Statistical Mechanics & Its Applications, 2016, 461(9):101-116.
[12] KOREN Y, BELL R, VOLINSKY C. Matrix factorization techniques for recommender systems[J]. Computer, 2009, 42(8):30-37.
[13] SALAKHUTDINOV R, MNIH A. Probabilistic matrix factorization[C]//International Conference on Neural Information Processing Systems. Vancouver, British Columbia, Canada:Curran Associates Inc. 2007:1257-1264.
[14] 谭学清, 蔡军, 罗琳. 基于改进的LSI标签语义检索书目系统[J]. 图书馆学研究, 2014,(11):67-72.TAN X Q, CAI J, LUO L. The tag semantic retrieval of bibliography system based on the improved LSI[J]. Research on Library Science, 2014,(11):67-72.
[15] CANTADOR I, BRUSILOVSKY P, KUFLIK T. Proceedings of the 2nd international workshop on information heterogeneity and fusion in recommender systems (HetRec 2011):27th October 2011, Chicago, IL, USA[J]. Soins Psychiatrie, 2011,(232):5-6.
[16] 卢露, 魏登月. 一种基于隐语义模型的协同过滤算法[J]. 微电子学与计算机, 2015,(2):73-75.LU L, WEI D Y. A collaborative filtering algorithm based on latent factor model[J]. Microelectronics & Computer, 2015,(2):73-75.
[1] 胡晓敏, 龙祖涛, 李敏. 基于用户分层的多目标推荐算法[J]. 广东工业大学学报, 2023, 40(01): 10-18.
[2] 谢光强, 许浩然, 李杨, 陈广福. 基于多智能体强化学习的社交网络舆情增强一致性方法[J]. 广东工业大学学报, 2022, 39(06): 36-43.
[3] 饶东宁, 王军星, 魏来, 王雅丽. 并行最小割算法及其在金融社交网络中的应用[J]. 广东工业大学学报, 2018, 35(02): 46-50.
[4] 胡惠成, 陈平华. 一种融合隐式信任关系的推荐算法[J]. 广东工业大学学报, 2017, 34(03): 43-48.
[5] 饶东宁, 温远丽, 魏来, 王雅丽. 基于Spark平台的社交网络在不同文化环境中的中心度加权算法[J]. 广东工业大学学报, 2017, 34(03): 15-20.
[6] 王晓彤. 基于PageRank的微博用户影响力度量[J]. 广东工业大学学报, 2016, 33(03): 49-54.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发