结合用户聚类和评分偏好的推荐算法 - 计算机应用研究 编辑部 - 《计算机应用研究》唯一官方网站

《计算机应用研究》|Application Research of Computers

结合用户聚类和评分偏好的推荐算法

Recommendation algorithm based on user clustering and rating preference

免费全文下载 (已被下载 次)  
获取PDF全文
作者 高茂庭,段元波
机构 上海海事大学 信息工程学院,上海 201306
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2018)08-2260-05
DOI 10.3969/j.issn.1001-3695.2018.08.005
摘要 针对推荐算法中用户评分矩阵维度高、计算量大的问题,为更加真实地反映用户本身评分偏好,提出一种结合用户聚类和评分偏好的推荐算法。先利用PCA降维和K-means聚类对用户评分矩阵进行预处理,在最近邻选取方法上,添加用户共同评分数量作为约束,利用用户和相似簇的相似度对相似簇内评分加权求和生成基本预测评分;再综合用户评分偏置和用户项目类型偏好,建立用户评分偏好模型;最后通过多元线性回归确定每部分的权重,生成最终的预测评分。对比实验结果表明,新算法能更真实地反映用户评分,有效减少计算量并提高推荐系统的预测准确率,更好地满足用户对于推荐系统的个性化需求。
关键词 协同过滤;降维;聚类;用户偏好;推荐系统
基金项目 国家自然科学基金资助项目(61202022)
本文URL http://www.arocmag.com/article/01-2018-08-005.html
英文标题 Recommendation algorithm based on user clustering and rating preference
作者英文名 Gao Maoting, Duan Yuanbo
机构英文名 CollegeofInformationEngineering,ShanghaiMaritimeUniversity,Shanghai201306,China
英文摘要 To solve the problem of high dimensionality and computational complexity of the user scoring matrix in recommendation system and reflected the user’s preference more realistically, this paper proposed a recommendation algorithm based on user clustering and user rating preference. Firstly, it used the PCA dimensionality reduction and K-means clustering to preprocess user rating matrix, used the number of user common rating items as the constraint in the nearest neighbor selection and used the similarity between user and similar cluster to sum weighted scores and generated a basic prediction rating. Secondly, it established the user scoring preference model by combining the user rating bias and the user item type preference. Finally, it used multiple linear regression to determine the weight of each component and obtained the final prediction score. The experimental results show that the new algorithm can reflect the user rating more accurately, reduces the computational complexity and improves the prediction accuracy of the recommendation system effectively, and meets the user’s personalized requirements for the recommendation system better.
英文关键词 collaborative filtering; dimensionality reduction; clustering; user preference; recommendation system
参考文献 查看稿件参考文献
  [1] Herlocker L, Konstan A, Borchers S A, et al. An algorithmic framework for performing collaborative filtering[C] // Proc of the 22nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York:ACM Press, 1999:230-237.
[2] Goldberg D, Nichols D, Oki B M, et al. Using collaborative filtering to weave an information tapestry[J] . Communications of the ACM, 1992, 35(12):61-70.
[3] Adomavicius G, Tuzhilin A. Toward the next generation of recommender systems:a survey of the state-of-the-art and possible extensions[J] . IEEE Trans on Knowledge & Data Engineering, 2005, 17(6):734-749.
[4] Goldberg K, Roeder T, Gupta D, et al. Eigentaste:a constant time collaborative filtering algorithm[J] . Information Retrieval Journal, 2001, 4(2):133-151.
[5] You Haipeng, Li Hui, Wang Yunmin, et al. An improved collaborative filtering recommendation algorithm combining item clustering and slope one scheme[C] //Lecture Notes in Engineering & Computer Science, vol 2215. 2015:313-316.
[6] Frémal S, Lecron F. Weighting strategies for a recommender system using item clustering based on genres[J] . Expert Systems with Applications, 2017, 77(7):105-113.
[7] Guo Guibing, Zhang Jie, Yorke-Smith N. Leveraging multiviews of trust and similarity to enhance clustering-based recommender systems[J] . Knowledge-Based Systems, 2015, 74(1):14-27.
[8] 邓爱林, 左子叶, 朱扬勇. 基于项目聚类的协同过滤推荐算法[J] . 小型微型计算机系统, 2004, 25(9):1665-1670.
[9] 黄创光, 印鉴, 汪静, 等. 不确定近邻的协同过滤推荐算法[J] . 计算机学报, 2010, 33(8):1369-1377.
[10] Herlocker J. Clustering items for collaborative filtering[C] // Proc of ACM SIGIR Workshop on Recommender Systems. New York:ACM Press, 1999.
[11] Beutel A, Beutel A, Ahmed A, et al. ACCAMS:additive co-clustering to approximate matrices succinctly[C] //Proc of the 24th International Conference on World Wide Web. Switzerland:International World Wide Web Conferences Steering Committee, 2014:119-129.
[12] MovieLens_100K[DB/OL] . https://grouplens. org/datasets/mo-vielens/.
[13] Forsati R, Barjasteh I, Masrour F, et al. PushTrust:an efficient recommendation algorithm by leveraging trust and distrust relations[C] //Proc of the 9th ACM Conference on Recommender Systems. New York:ACM Press, 2015:51-58.
[14] Sarwar B, Karypis G, Konstan J, et al. Item-based collaborative filtering recommendation algorithms[C] //Proc of International Conference on World Wide Web. New York:ACM Press, 2001:285-295.
[15] Altingovde I S, Subakan  N, Ulusoy . Cluster searching strategies for collaborative recommendation systems[J] . Information Processing & Management, 2013, 49(3):688-697.
[16] 王茜, 杨莉云, 杨德礼. 面向用户偏好的属性值评分分布协同过滤算法[J] . 系统工程学报, 2010, 25(4):131-138.
[17] 郭均鹏, 赵梦楠. 面向在线社区用户的群体推荐算法研究[J] . 计算机应用研究, 2014, 31(3):696-699.
[18] 崔春生. 推荐系统中显式评分输入的用户聚类方法研究[J] . 计算机应用研究, 2011, 28(8):2856-2858.
[19] 范波, 程久军. 用户间多相似度协同过滤推荐算法[J] . 计算机科学, 2012, 39(1):23-26.
[20] 黄震华, 张佳雯, 田春岐, 等. 基于排序学习的推荐算法研究综述[J] . 软件学报, 2016, 27(3):691-713.
收稿日期 2017/4/10
修回日期 2017/5/15
页码 2260-2264
中图分类号 TP301.6
文献标志码 A