基于共享知识模型的跨领域推荐算法

doi:10.3969/j.issn.0372-2112.2018.08.020

电子学报 ›› 2018, Vol. 46 ›› Issue (8): 1947-1953.DOI: 10.3969/j.issn.0372-2112.2018.08.020

基于共享知识模型的跨领域推荐算法

李林峰, 刘真, 魏港明, 任爽, 葛梦凡

北京交通大学计算机与信息技术学院, 北京 100044

收稿日期:2017-07-26 修回日期:2017-12-17 出版日期:2018-08-25
- 通讯作者:
- 刘真
- 作者简介:
- 李林峰 女,1993年生于河南唐河.北京交通大学计算机科学与技术学院研究生.研究方向为大数据、推荐系统.
- 基金资助:
- 科技部国家重点研发计划 (No.2016YFB1200100）; 国家自然科学基金 (No.61202429，No.61763031）; 中央高校基本科研业务费专项 (No.2017JBM024）

Cross-Domain Recommendation Algorithm Based on Sharing Knowledge Pattern

LI Lin-feng, LIU Zhen, WEI Gang-ming, REN Shuang, GE Meng-fan

School of Computer and Information Technology, Beijing Jiaotong University, Beijing 100044, China

Received:2017-07-26 Revised:2017-12-17 Online:2018-08-25 Published:2018-08-25

摘要/Abstract

摘要： 互联网的普及使得大量信息不断累积，推荐系统作为解决信息过载的有效手段，能够帮助人们迅速准确地筛选出感兴趣的内容.但是由于用户项目评分数据过于稀疏，新用户或新商品存在"冷启动"问题，使得传统的推荐算法计算复杂性过高、准确性较低.考虑到用户会在互联网不同领域使用各类应用，在不同领域积累了大量行为数据和评价信息.而从用户群体的角度来说，在不同领域间存在着用户群体的偏好相似性，因此如果通过在不同领域中共享代表偏好的知识模型，将有助于提升在新领域推荐的准确性，解决冷启动问题.本文提出了基于共享知识模型的跨领域推荐算法SKP （Sharing Knowledge Pattern），通过对各个领域中用户-项目的评分矩阵分解，得到用户的潜在特征矩阵和项目的潜在特征矩阵，对用户和项目的潜在特征分别聚类，得到了用户分组对项目分组的评分知识模型，最终利用目标领域的个性知识模型和各个领域的共性知识模型来得出推荐结果.本文对三个不同领域的数据集进行了分析和划分，并在物理集群环境下进行了实验.结果表明，通过利用数据稠密的辅助领域数据，本文提出的SKP算法与已有的单领域算法、跨领域算法相比，具有更高的准确率和更低的RMSE值.

关键词: 跨领域, 推荐算法, 冷启动, 潜在因子, 知识模型

Abstract: With the popularity of the Internet and the accumulation of large amounts of data,recommendation system,as an effective means to solve the problem of information overload,can help people quickly select what they are interested in.Because of the sparse user-item rating data,and the cold start problem of new users or new items,traditional recommendation algorithm has the shortcoming of high complexity,low accuracy.Considering the accumulated users behavior or rating data across different domains can have the same preferences,we can share the knowledge pattern among different domains.Based on the matrix factorizationof user-itemrating data in different domains,we can obtain the latent feature matrix of users and items respectively.Considering the user group preference,the latent features of users and items are clustered separately as the domainknowledge pattern.Moreover,By clustering the cross-domain knowledge patterns,we can get shared common knowledge pattern.With the domain knowledge pattern and the shared common knowledge pattern,we can make the finalrecommendation.Based on the above consideration,this paper proposes the SKP (Sharing Knowledge Pattern)algorithm.And the SKP is realized in a parallel manner.Experimentsare carried out in the physical cluster environment.By exploiting three different datasets,the results show that the SKP algorithm has better recommendation accuracy and lower RMSE values compared with the existing single-domain algorithm and other cross-domainalgorithms.

Key words: cross-domain, recommendation algorithm, cold start, latent factor, knowledge pattern

中图分类号:

TN95

李林峰, 刘真, 魏港明, 任爽, 葛梦凡. 基于共享知识模型的跨领域推荐算法[J]. 电子学报, 2018, 46(8): 1947-1953.

LI Lin-feng, LIU Zhen, WEI Gang-ming, REN Shuang, GE Meng-fan. Cross-Domain Recommendation Algorithm Based on Sharing Knowledge Pattern[J]. Acta Electronica Sinica, 2018, 46(8): 1947-1953.

参考文献

[1] Mirbakhsh N,Ling C X.Improving top-n recommendation for cold-start users via cross-domain information[J].ACM Transactions on Knowledge Discovery from Data (TKDD),2015,9(4):http://dx.doi.org/10.1145/2724720.
[2] 张亮,柏林森,周涛.基于跨电商行为的交叉推荐算法[J].电子科技大学学报,2013(1):154-160. Zhang Liang,Bai Linsen,Zhou Tao.Crossover recommendation algorithm based on cross-merchants behavior[J].Journal of University of Electronic Science and Technology of China,2013(1):154-160.(in Chinese)
[3] Elkahky A M,Song Y,He X.A multi-view deep learning approach for cross domain user modeling in recommendation systems[A].Proceedings of the 24th International Conference on World Wide Web[C].International World Wide Web Conferences Steering Committee,2015.278-288.
[4] Fernández-Tobías I,Tomeo P,Cantador I,et al.Accuracy and diversity in cross-domain recommendations for cold-start users with positive-only feedback[A].Proceedings of the 10th ACM Conference on Recommender Systems[C].ACM,2016.119-122.
[5] Chen L,Zheng J,Gao M,et al.TLRec:transfer learning for cross-domain recommendation[A].Big Knowledge (ICBK),2017 IEEE International Conference on[C].IEEE,2017.167-172.
[6] Taneja A,Arora A.Cross domain recommendation using multidimensional tensor factorization[J].Expert Systems with Applications,2018,92:304-316.
[7] Zheng L,Noroozi V,Yu P S.Joint deep modeling of users and items using reviews for recommendation[A].Proceedings of the Tenth ACM International Conference on Web Search and Data Mining[C].ACM,2017.425-434.
[8] 王兴茂,张兴明,吴毅涛,潘俊池.基于启发式聚类模型和类别相似度的协同过滤推荐算法[J].电子学报,2016,44(7):1708-1713. Wang XingMao,Zhang Xingming,Wu Yitao,Pan Junchi.Collaborative filtering recommendation algorithm based on heuristic clustering model and class similarity[J].Acta Electronica Sinica,2016,44(7):1708-1713.(in Chinese)
[9] Xu Z,Zhang F,Wang W,et al.Exploiting trust and usage context for cross-domain recommendation[J].IEEE Access,2016,4:2398-2407.
[10] Singh A P,Gordon G J.Relational learning via collective matrix factorization[A].ACM SIGKDD International Conference on Knowledge Discovery and Data Mining[C].ACM,2008.650-658.
[11] Jamali M,Lakshmanan L.HeteroMF:recommendation in heterogeneous information networks using context dependent factor models[A].International Conference on World Wide Web[C].ACM,2013.643-654.
[12] Shi J,Long M,Liu Q,et al.Twin Bridge Transfer Learning for Sparse Collaborative Filtering[M].Advances in Knowledge Discovery and Data Mining.2013.496-507.
[13] Yan Z,Wei L,Lu Y,et al.You are what apps you use:Transfer Learning for Personalized Content and Ad Recommendation[A].Proceedings of the Eleventh ACM Conference on Recommender Systems[C].ACM,2017.350-350.
[14] Jiang M,Cui P,Chen X,et al.Social recommendation with cross-domain transferable knowledge[J].IEEE Transactions on Knowledge & Data Engineering,2015,27(11):3084-3097.
[15] Li B,Yang Q,Xue X.Can Movies and books collaborate cross-domain collaborative filtering for sparsity reduction[A].IJCAI 2009,Proceedings of the International Joint Conference on Artificial Intelligence[C].Pasadena,California,Usa:DBLP,2009.2052-2057.
[16] Li B,Yang Q,Xue X.Transfer learning for collaborative filtering via a rating-matrix generative model[A].International Conference on Machine Learning,ICML 2009[C].Montreal,Quebec,Canada:DBLP,2009.78.
[17] Fernández-Tobías I,Tomeo P,Cantador I,et al.Accuracy and diversity in cross-domain recommendations for cold-start users with positive-only feedback[A].Proceedings of the 10th ACM Conference on Recommender Systems[C].ACM,2016.119-122.
[18] Pan W.A survey of transfer learning for collaborative recommendation with auxiliary data[J].Neurocomputing,2016,177:447-453.
[19] Gao S,Luo H,Chen D,et al.Cross-domain recommendation via cluster-level latent factor model[A].Joint European Conference on Machine Learning and Knowledge Discovery in Databases[C].Berlin,Heidelberg:Springer,2013.161-176.
[20] Berkovsky S,Kuflik T,Ricci F.Mediation of user models for enhanced personalization in recommender systems[J].User Modeling and User-Adapted Interaction,2008,18(3):245-286.
[21] Hoyer P O.Non-negative matrix factorization with sparseness constraints[J].Journal of Machine Learning Research,2004,5(1):1457-1469.

基于共享知识模型的跨领域推荐算法

Cross-Domain Recommendation Algorithm Based on Sharing Knowledge Pattern

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

编辑推荐

Metrics

[1]	郭景峰, 朱晓松, 李爽. 直播电视推荐系统的评分预测算法研究[J]. 电子学报, 2020, 48(9): 1735-1740.
[2]	杨洁, 朱咸军, 周献中, 柳毅. 基于混杂社会网络的个性化Web服务推荐方法[J]. 电子学报, 2020, 48(2): 341-349.
[3]	刘真, 田靖玉, 苑宝鑫, 孙永奇. 基于知识聚合和迁移的跨领域推荐算法[J]. 电子学报, 2020, 48(10): 1928-1932.
[4]	俞春花, 刘学军, 李斌, 章玮. 基于上下文相似度和社会网络的移动服务推荐方法[J]. 电子学报, 2017, 45(6): 1530-1536.
[5]	黄震华, 张佳雯, 张波, 喻剑, 向阳, 黄德双. 语义推荐算法研究综述[J]. 电子学报, 2016, 44(9): 2262-2275.
[6]	张维玉, 吴斌, 耿玉水, 朱江. 基于协同矩阵分解的评分与信任联合预测[J]. 电子学报, 2016, 44(7): 1581-1586.
[7]	王兴茂, 张兴明, 吴毅涛, 潘俊池. 基于启发式聚类模型和类别相似度的协同过滤推荐算法[J]. 电子学报, 2016, 44(7): 1708-1713.
[8]	印桂生, 张亚楠, 董宇欣, 韩启龙. 基于受限信任关系和概率分解矩阵的推荐[J]. 电子学报, 2014, 42(5): 904-911.