# 基于电子商务的协同过滤推荐算法综述 A Research Summary of Collaborative Filtering Recommendation Algorithm Based on E-Commerce

• 全文下载: PDF(444KB)    PP.1-5   DOI: 10.12677/ECL.2019.81001
• 下载量: 213  浏览量: 447

The collaborative filtering recommendation algorithm is the most abundant and most used recommendation algorithm in the current e-commerce recommendation system. This paper first introduces three different collaborative filtering recommendation algorithms in detail and compares their respective advantages and disadvantages. Then, the current research status of data sparseness, cold start and scalability of collaborative filtering recommendation algorithm is summarized. And this paper analyzes the merits and deficiencies of the existing methods. Finally, the research hotspot of future collaborative filtering algorithm is proposed, which provides a reference for the future development of collaborative filtering algorithm.

1. 引言

2. 协同过滤推荐算法

2.1. 基于用户的协同过滤推荐算法

GroupLens系统 [3] 所使用的就是基于用户的协同过滤算法。它的基本假设就是“对于项目评分相似度高的用户其兴趣或偏好也是相似的即对同一个项目的喜好程度是一致的”。该算法让用户根据自己的兴趣对购买过的项目进行打分，然后分析比较该用户与其他用户的评分数据，进而计算他们评分之间的相似度，相似度越高表明他们偏好也就越相似。最后利用找到的与目标用户相似的用户的评分来预测目标用户对其未评分项目的喜好程度完成推荐。皮尔森系数与余弦相似度的计算公式分别如下：

${\rho }_{x,y}=\frac{E\left(XY\right)-E\left(X\right)E\left(Y\right)}{\sqrt{E\left({X}^{2}\right)-{E}^{2}\left(X\right)}\sqrt{E\left({Y}^{2}\right)-{E}^{2}\left(Y\right)}}$

$\mathrm{cos}\theta =\frac{{\sum }_{n}^{1}\left({A}_{i}×{B}_{i}\right)}{\sqrt{{\sum }_{n}^{1}\left({A}_{i}^{2}\right)}×\sqrt{{\sum }_{n}^{1}\left({B}_{i}^{2}\right)}}$

2.2. 基于项目的协同过滤推荐算法

2.3. 基于模型的协同过滤推荐算法

2.4. 三种协同过滤推荐算法比较

Table 1. Comparison of advantages and disadvantages of three collaborative filtering algorithms and application scenarios

3. 协同过滤算法存在的问题及解决方法

3.1. 数据稀疏性问题

3.2. 冷启动问题

3.3. 可拓展性问题

4. 总结与展望

 [1] Herlocker, J., Konstan, J. and Riedl, J. (2000) Explaining Collaborative Filtering Recommendations. Proceedings of the 2000 ACM Conference on Computer Supported Cooperative Work, ACM Press, New York, 241-250. https://doi.org/10.1145/358916.358995 [2] Goldberg, D. and Nicols. D. (1992) Using Collaborative Filtering to Weave an Information Tapstry, Communications of the ACM, 35, 61-70. https://doi.org/10.1145/138859.138867 [3] Resnick, P., et al. (1994) GroupLens: An Open Architecture for Collaborative Filtering of Netnews. Proceedings of ACM Conference on Computer Supported Cooperative Work, CSCW 1994, Chapel Hill, 175-186. https://doi.org/10.1145/192844.192905 [4] Breese, J.S., Heckerman, D. and Kadie, C. (1998) Empirical Analysis of Predictive Algorithms for Collaborative Filtering. Proceedings of the 14th Conference on Uncertainty in Artificial Intelligence, San Francisco, 24-26 July 1998, 43-52. [5] Sarwar, B., Karypis, G., Konstan, J. and Riedl, J. (2001) Item-Based Collaborative Filtering Recommendation Algorithms. Proceeding of the 10th international conference on World Wide Web, Hong Kong, 1-5 May 2001. [6] 冷亚军, 陆青, 梁昌勇. 协同过滤推荐技术综述[J]. 模式识别与人工智能, 2014, 27(8): 720-734. [7] 孙小华. 协同过滤推荐系统的稀疏性与冷启动问题研究[D]: [博士学位论文]. 杭州: 浙江大学, 2006. [8] 焦明海, 陈晓芳, 陈旭, 等. 基于贝叶斯网络认知反馈的协同过滤推荐[J]. 控制工程, 2017, 24(7): 1310-1317. [9] 周朴雄, 张兵荣, 赵龙文. 基于BP神经网络的情境化信息推荐服务研究[J]. 情报科学, 2016, 36(3): 71-75. [10] 王建芳, 李骁, 武文琪, 等. 一种SVD和信任因子相结合的协同过滤算法[J]. 小型微型计算机系统, 2017, 38(6): 1290-1293. [11] Funk, S. (2006) Netflix Update: Try This at Home[EB/OL]. http://sifter.org/~simon/journal/20061211.html, 2017-9-15 [12] Koren, Y. (2008) Factorization Meets the Neighborhood: A Multifaceted Collaborative Filtering Model. Proceedings of International Conference on Knowledge Discovery and Data Mining, 426-434. [13] Kim, H.N., Alkhaldi, A., Saddik, A.E., et al. (2011) Collaborative User Modeling with User-Generated Tags for Social Recommender Systems. Expert Systems with Applications, 38, 8488-8496. https://doi.org/10.1016/j.eswa.2011.01.048 [14] 李聪, 梁昌勇. 基于n序访问解析逻辑的协同过滤冷启动消除方法[J]. 系统工程理论与实践, 2012(7): 1537-1545. [15] 郭晓波, 赵书良, 牛东攀. 一种解决稀疏数据和冷启动问题的组合推荐方法[J]. 中国科学技术大学学报, 2015, 45(10): 804-812. [16] 于洪, 李俊华. 一种解决新项目冷启动问题的推荐算法[J]. 软件学报, 2015, 26(6): 1395-1408. [17] 吴湖, 王永吉, 王哲, 等. 两阶段联合聚类协同过滤算法[J]. 软件学报, 2010, 21(S): 1042-1054.