基于潜在因子模型的跨领域信息推荐算法

doi:10.11959/j.issn.1000-0801.2015188

电信科学 ›› 2015, Vol. 31 ›› Issue (7): 75-79.doi: 10.11959/j.issn.1000-0801.2015188

基于潜在因子模型的跨领域信息推荐算法

高升,任思婷,郭军

北京邮电大学北京 100876

出版日期:2015-08-21 发布日期:2015-08-21
基金资助:
国家自然科学基金资助项目;>国家自然科学基金资助项目;教育部—中国移动科研基金资助项目

Cross-Domain Recommendation Algorithm Based on Latent Factor Model

Sheng Gao,Siting Ren,Jun Guo

Beijing University of Posts and Telecommunications，Beijing 100876，China

Online:2015-08-21 Published:2015-08-21
Supported by:
The National Science Foundation of China;The National Science Foundation of China;Ministry of Education-China Mobile Research Foundation

摘要/Abstract

摘要：

互联网环境下，不同领域中多源异构信息对象的交互融合使用户面临大数据环境下的信息选择困境，传统的信息推荐算法已很难适应跨领域的信息推荐服务。综合分析了不同领域内用户对信息对象的评价数据，基于潜在因子模型抽取了不同领域中某一用户聚类集合对某一信息对象聚类集合评分模式的跨领域共性特征和单领域个性特征，进而通过传递、共享跨域共性特征信息的方式缓解了目标领域的数据稀疏性问题，提高了跨域信息推荐的准确度。

关键词: 跨领域信息推荐, 潜在因子模型, 用户评价模式

Abstract:

In the internet environment，the combining of multi-source heterogeneous information objects in different areas makes users face information selection dilemma problem in big data environment.It has been very difficulty for traditional information recommendation algorithms to adapt to the interdisciplinary information recommendation service.The evaluation model from a user clustering set to an information object clustering set has common characteristics of cross-domain and personality characteristics of single domain.By analyzing the evaluation data from users to information objects in different areas，these characteristics were extracted based on latent factor model.Then by transmitting and sharing the common characteristics of cross-domain，the data sparseness problem of target field was alleviated，which could improve the accuracy of cross-domain information recommendation.

Key words: cross-domain recommendation, latent factor model, user rating pattern

高升,任思婷,郭军. 基于潜在因子模型的跨领域信息推荐算法[J]. 电信科学, 2015, 31(7): 75-79.

Sheng Gao,Siting Ren,Jun Guo. Cross-Domain Recommendation Algorithm Based on Latent Factor Model[J]. Telecommunications Science, 2015, 31(7): 75-79.

图/表 4

图1

表1

表2

表3

参考文献 18

1	Adomavicius G ， Tuzhilin A . Toward the next generation of recommender systems：a survey of the state-of-the-art and possible extensions. IEEE Transactions on Knowledge and Data Engineering（TKDE）， 2005，17（6）： 734～749
2	Koren Y . Collaborative filtering with temporal dynamics. Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining， Paris，France， 2009： 447～455
3	Jordan M I ， Blei D M ， Ng A Y ，et al. Latent dirichlet allocation. Journal of Machine Learning Research， 2003（3）： 993～1022
4	Hofmann T . Probabilistic latent semantic indexing. Proceedings of the 21th Annual International SIGIR Conference on Research ＆ Development in Information Retrieval， Berkeley，CA，USA， 1999： 457～465
5	Hofmann T . Collaborative filtering via gaussian probabilistic latent semantic analysis. Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval， Toronto，Canada， 2003： 259～266
6	Bishop C M . Pattern Recognition and Machine Learning. New York： Springer-Verlag， 2006
7	Lee D D ， Seung H S . Learning the parts of objects by nonnegative matrix factorization. Nature， 1999，401（6755）： 788～791
8	Olshausen B A ， Field D J . Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature， 1996（381）： 607～609
9	Gao S ， Luo H ， Chen D ，et al. Cross-domain recommendation via cluster-level latent factor model. Proceedings of European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Database， Prague，Czech Republic， 2013： 161～176
10	Pan W ， Liu N N ， Xiang E W ，et al. Transfer learning to predict missing ratings via heterogeneous user feedbacks. Proceedings of the 22th International Joint Conference on Artificial Intelligence， Barcelona，Spain， 2011： 2318～2323
11	Li B ， Yang Q ， Xue X . Transfer learning for collaborative filtering via a rating-matrix generative model. Proceedings of the 26th International Conference on Machine Learning， Montreal，Canada， 2009： 617～624
12	Singh A P ， Gordon G J . Relational learning via collective matrix factorization. Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining， Las Vegas，Nevada，USA， 2008： 650～658
13	Berkovsky S ， Kuflik T ， Ricci F . Mediation of user models for enhanced personalization in recommender systems. User Modeling and User-Adapted Interaction， 2008，18（3）： 345～352
14	Zhang Y ， Koren J . Relational learning via collective matrix factorization. Proceedings of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval， New York，USA， 2007： 47～54
15	Lee D D ， Seung H S . Algorithms for non-negative matrix factorization. Advances in Neural Information Processing Systems， 2001，13（3）： 556～562
16	Luo S ， Rong J . Flexible mixture model for collaborative filtering. Proceedings of the 20th International Conference on Machine Learning， Washington DC，USA， 2003（3）： 704～711
17	Baeza-Yates R ， Ribeiro-Neto B . Modern information retrieval. Facet Publishing， 1999，20（19）： 305～306
18	Hu Y F ， Koren Y ， Volinsky C . Collaborative filtering for implicit feedback datasets. Proceedings of IEEE 8th Conference on Data Mining， Pisa，Italy， 2008： 263～272

数据集	算法	Given5	Given10	Given15
Book-Crossing	NMF	0.657 5	0.637 5	0.632 1
	FMM	0.645 1	0.619 6	0.612 5
	RMGM	0.637 8	0.611 5	0.609 2
	PCLF	0.625 2	0.599 4	0.596 9
EachMovie	NMF	0.934 5	0.886 1	0.879 9
	FMM	0.913 2	0.883 1	0.877 1
	RMGM	0.902 1	0.874 3	0.863 7
	PCLF	0.883 8	0.867 7	0.853 3

数据集	算法	Given5	Given10	Given15
Book-Crossing	NMF	0.657 5	0.637 5	0.630 1
	FMM	0.645 1	0.619 6	0.612 5
	RMGM	0.636 4	0.605 4	0.599 4
	PCLF	0.628 4	0.599 0	0.595 0
MovieLens	NMF	0.836 5	0.801 6	0.793 3
	FMM	0.805 2	0.783 8	0.772 1
	RMGM	0.803 9	0.776 9	0.769 2
	PCLF	0.792 2	0.772 9	0.765 3

数据集	算法	Given5	Given10	Given15
MovieLens	NMF	0.836 5	0.801 6	0.793 3
	FMM	0.805 2	0.783 8	0.772 1
	RMGM	0.798 5	0.773 6	0.767 7
	PCLF	0.787 8	0.772 4	0.763 9
EachMovie	NMF	0.934 5	0.886 1	0.879 9
	FMM	0.913 2	0.883 1	0.877 1
	RMGM	0.901 7	0.873 3	0.860 1
	PCLF	0.882 0	0.865 8	0.849 8

基于潜在因子模型的跨领域信息推荐算法

Cross-Domain Recommendation Algorithm Based on Latent Factor Model

在线阅读

PDF下载

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 4

参考文献 18

相关文章 0

Metrics

推荐阅读 0