OU Ruyue,TAO Hongcai.A Multi-layer Personalized Recommendation Algorithm Combining PageRank and PersonalRank[J].Journal of Chengdu University of Information Technology,2021,36(03):305-310.[doi:10.16836/j.cnki.jcuit.2021.03.011]
一种融合PageRank和PersonalRank的多层个性化推荐算法
- Title:
- A Multi-layer Personalized Recommendation Algorithm Combining PageRank and PersonalRank
- 文章编号:
- 2096-1618(2021)03-0305-06
- 关键词:
- 推荐系统; 多层推荐; PageRank; PersonalRank; 图模型
- 分类号:
- TP311.5
- 文献标志码:
- A
- 摘要:
- 传统的推荐系统只能实现一种类型的实体推荐,为解决一次性进行多种类型实体即多层推荐的问题,提出一种融合PageRank和PersonalRank的多层个性化推荐算法。利用图数据模型中的顶点描述实体,边描述实体间关联关系的这种特性,在图中将用户作为第一层实体即起始点,而将用户的历史行为(如评论过的电影)作为第二层实体,根据第二层实体依次给用户推荐第三层、第四层直到第N层的实体列表。通过爬虫爬取豆瓣网电影获取数据集,实验结果表明该模型具有多层推荐的效果,并较PersonalRank算法有更高的准确率和召回率。
- Abstract:
- The traditional recommendation system can only realize the recommendation of one type of entity. In order to solve the problem of multiple types of entities at one time, that is, multi-layer recommendation, a multi-layer personalized recommendation algorithm combining PageRank and PersonalRank algorithms is proposed. Firstly, it utilizes the characteristics of using vertices to describe entities in the graph data model, and describes the relationship among entities by edges. Then, it takes users as the first-level entities in the graph, i.e.,the starting point, and the historical behaviors(e.g., reviewed movies)left by users as the second-level entities. Further, according to the second layer the third layer is recommended to the user in turn, and the fourth layer up to the Nth layer of the entity list is also recommended. The experiment on the data set obtained by crawling Douban movies shows that the model has a multi-layer recommendation effect, and has a higher accuracy and recall rate than the original PersonalRank algorithm.
参考文献/References:
[1] Xuanhui Wang,Tao Tao,Jiantao Sun,et al.DirichletRank:Solving the Zero-One Gap Problem of PageRank[J].ACM Transactions on Information Systems,2008,26(2):1-66.
[2] 库珊,刘钊.基于PageRank与HITS的改进算法的网页排名优化[J]. 武汉科技大学学报,2019,42(2):155-160.
[3] 戴炳荣,姜胜明,李顿伟,等.基于改进PageRank算法的跨链公证人机制评价模型[J/OL].计算机工程,https://doi.org/10.19678/j.issn.1000-3428.0056460,2020-05-24.
[4] Bollen J,Rodriquez M A,Van De Sompel H.Journal status[J].Scientometrics,2006,69(3):669-687.
[5] 马凤.基于PageRank算法的期刊影响力研究[J].情报杂志,2014,33(12):103-108.
[6] Chen P,Xie H,Maslov S,Redner S.Finding scientific gems with Google’s PageRank algorithm[J].Journal of Informetrics,2007,1(1):8-15.
[7] 苏成,Hee-Sop KIM.基于PageRank,HITS和SALSA算法的学术论文评价[J].情报杂志,2015,34(6):48-54.
[8] Ding Y,Yan E,Frazho A,Caverlee J.PageRank for ranking authors in co-citation netw orks[J]. Journal of the American Society for Information Science and Technology,2009,60(11):2229-2243.
[9] Fiala D.Suborganizations of institutions in Library and Information Science Journals[J].Information, 2013,4(4):351-366.
[10] 邱苓芸,王铭,赵卫东.PageRank算法改进研究[J].软件导刊,2017,16(2):74-76.
[11] 刘清,王帆,冯亮,等.高效图推荐算法应用研究[J].软件导刊,2019,18(8):49-51.
[12] 吴迪,周利娟,林鸿飞.基于随机游走的就业推荐系统研究与实现[J].广西师范大学学报(自然科学版),2011(1):179-185.
[13] 杨华.基于图的商品推荐算法研究[D].南昌:江西师范大学,2017.
[14] 滕传志,赵月旭.基于随机森林-马尔可夫用户冷启动推荐系统[J].计算机工程与设计,2020,41(11):3094-3098.
相似文献/References:
[1]谢海迪,周 云,李彤岩.一种基于社交网络的非负矩阵分解算法[J].成都信息工程大学学报,2024,39(01):50.[doi:10.16836/j.cnki.jcuit.2024.01.009]
XIE Haidi,ZHOU Yun,LI Tongyan.A Non-negative Matrix Decomposition Algorithm based on Social Networks[J].Journal of Chengdu University of Information Technology,2024,39(03):50.[doi:10.16836/j.cnki.jcuit.2024.01.009]
备注/Memo
收稿日期:2020-03-28
基金项目:国家自然科学基金资助项目(61806170)