[1]秦倩倩,张志军,袁卫华,等.融合元路径与改进协同注意力的推荐研究[J].计算机技术与发展,2022,32(12):150-158.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 023]
　QIN Qian-qian,ZHANG Zhi-jun,YUAN Wei-hua,et al.Research on Recommendation Fusing Meta-Path and Improved Collaborative Attention[J].,2022,32(12):150-158.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 023]
点击复制

融合元路径与改进协同注意力的推荐研究()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年12期
页码:
150-158
栏目:
人工智能
出版日期:
2022-12-10

文章信息/Info

Title:
Research on Recommendation Fusing Meta-Path and Improved Collaborative Attention
文章编号:
1673-629X(2022)12-0150-09
作者:
秦倩倩; 张志军; 袁卫华; 桑 胜; 胡海星; 王宜贵*
山东建筑大学 计算机科学与技术学院,山东 济南 250101
Author(s):
QIN Qian-qian; ZHANG Zhi-jun; YUAN Wei-hua; SANG Sheng; HU Hai-xing; WANG Yi-gui*
School of Computer Science and Technology,Shandong Jianzhu University,Jinan 250101,China
关键词:
异质信息网络; 推荐系统; 协同注意力机制; 元路径; 最大池化
Keywords:
heterogeneous information network; recommendation system; collaborative attention mechanism; meta-path; max pooling
分类号:
TP391
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 12. 023
摘要:
现有的基于异质信息网络的推荐方法通过最大池化操作提取元路径语义信息时,未考虑元路径语义的整体性特征及特征冗余问题的影响。 针对这些问题,提出了一种融合元路径与改进协同注意力的推荐模型 MICA( Research on Rec鄄ommendation Fusing Meta-Path and Improved Collaborative Attention) 。 该模型通过基于 k-最大池化的协同注意力机制更深层次地学习用户和项目的协同注意力嵌入表达,缓解了最大池化操作在协同注意力机制中带来的特征丢失问题;基于提取的路径序列显著性特征和完整性特征通过注意力机制学习元路径的重要特征,得到元路径上下文嵌入表达,保留有效的元路径语义信息;利用注意力机制学习融合后的用户、项目协同注意力嵌入表达和元路径上下文嵌入表达,以减少冗余信息,实现 top-N 推荐。 在两个真实数据集上的实验结果表明,MICA 在三个评价指标上均优于其他模型,能够更好地提高推荐性能,有效地提取特征信息,缓解元路径特征提取不充分等问题。
Abstract:
When the existing recommendation methods based on heterogeneous information network extract meta - path semanticinformation by maximum pooling operation,the influence of the overall characteristics of meta - path semantics and feature redundancycannot be taken into account. To solve these problems,we propose a recommendation model named MICA,which fuses meta-path andimproved collaborative attention mechanism. Firstly,MICA uses a collaborative attention mechanism based on k-max pooling to learn thecollaborative attention embedding representations of users and items in a deeper level,respectively,which alleviates the problem of featurelosses caused by max pooling in the collaborative attention mechanism. Secondly,based on the significant feature and integral feature,　themeta-path context embedding representation is obtained through the attention mechanism,a mechanism to learn important feature of meta-path,to retain effective meta - path information. Lastly, to reduce redundant information, the user and item collaborative attentionembedding representations and meta-path context embedding representations are fused to input into attention mechanism for top-N recommendations. Experiments on two real datasets show that MICA is superior to other state - of - the - art models in three evaluationindices,which better improves the recommendation performance. The model can effectively extract feature information and alleviate theproblem of insufficient feature extraction of meta-path.

相似文献/References:

[1]邵延振 蒙韧 袁鼎荣 李新友.基于Web结构分区的协同过滤推荐算法研究[J].计算机技术与发展,2010,(06):67.
　SHAO Yan-zhen,MENG Ren,YUAN Ding-rong,et al.Collaborative Filtering Recommendation Algorithm Research Based on Web Blocks[J].,2010,(12):67.
[2]曹毅 贺卫红.基于内容过滤的电子商务推荐系统研究[J].计算机技术与发展,2009,(06):182.
　CAO Yi,HE Wei-hong.Research on E- Commerce Recommender System Based on Content - Based Filtering[J].,2009,(12):182.
[3]赵鹏 蔡庆生 王清毅.一种用于文章推荐系统中的用户模型表示方法[J].计算机技术与发展,2007,(01):4.
　ZHAO Peng,CAI Qing-sheng,WANG Qing-yi.A Novel Representation of User Profile in Document Recommendation System[J].,2007,(12):4.
[4]高静 应吉康.基于人工免疫系统的推荐系统[J].计算机技术与发展,2007,(05):180.
　GAO Jing,YING Ji-kang.A Recommendation System Based on Artificial Immune System[J].,2007,(12):180.
[5]游文 叶水生.电子商务推荐系统中的协同过滤推荐[J].计算机技术与发展,2006,(09):70.
　YOU Wen,YE Shui-sheng.A Survey of Collaborative Filtering Algorithm Applied in E- commerce Recommender System[J].,2006,(12):70.
[6]华铨平.基于FNN的家纺产品个性化推荐系统的研究[J].计算机技术与发展,2011,(09):183.
　HUA Quan-ping.Research on Personalized Recommendation System of Textile Products Used by Family Based on FNN[J].,2011,(12):183.
[7]马言春 彭志平.基于市场机制的云服务管理研究[J].计算机技术与发展,2012,(03):214.
　MA Yan-chun,PENG Zhi-ping.Research on Cloud Service Management Based on Market Mechanism[J].,2012,(12):214.
[8]刘厚良.网络协同戏剧中个性化戏剧资源推荐系统[J].计算机技术与发展,2012,(08):25.
　LIU Hou-liang.Personalized Drama Resource Recommendation System in Cooperative Play[J].,2012,(12):25.
[9]陆晓敏 崇志宏 陈国庆.基于本体的商品推荐方法[J].计算机技术与发展,2012,(10):10.
　LU Xiao-min,CHONG Zhi-hong,CHEN Guo-qing.Product Recommending Method Based on Ontology[J].,2012,(12):10.
[10]范虎,花伟伟.协同过滤推荐算法的研究与改进[J].计算机技术与发展,2013,(09):66.
　FAN Hu[],HUA Wei-wei[].Research and Improvement of Collaborative Filtering Recommendation Algorithm[J].,2013,(12):66.

更新日期/Last Update: 2022-12-10