基于双头自编码器的评论主题感知推荐模型

刘树栋,李震,郝熙平,陈旭

PDF(3456 KB)
PDF(3456 KB)
中文信息学报 ›› 2024, Vol. 38 ›› Issue (9) : 146-166.
信息检索

基于双头自编码器的评论主题感知推荐模型

  • 刘树栋1,2,李震1,2,郝熙平1,2,陈旭1,2
作者信息 +

Two-headed Autoencoder Based Topic-aware Recommendation Model for Reviews

  • LIU Shudong1,2, LI Zhen1,2, HAO Xiping1,2, CHEN Xu1,2
Author information +
History +

摘要

近年来,推荐系统逐渐成为电子商务、在线流媒体、新闻资讯等各大互联网平台不可缺少的关键技术。以协同过滤技术为代表的推荐系统主要研究用户-项目评分数据,但此类方法常常面临新加入用户与用户交互次数少而导致的冷启动问题和数据稀疏问题。为解决上述问题,研究人员将用户和项目的上下文信息引入到协同过滤推荐系统中,丰富用户与项目表示。随着文本挖掘技术的发展,有研究发现用户对项目的评论文本不仅能够体现项目在不同方面的语义特征,也可以弥补用户-项目评分矩阵不能全面地反映用户语义偏好的局限,故可以将其应用到推荐系统中缓解数据稀疏性和冷启动问题。由于文本数据和用户-项目评分数据在用户偏好表示上存在差异,目前大多数模型在用户表示学习方面没有进行深层次的多次融合,为此,该文提出一种基于双头自编码器的评论感知推荐模型(Review Topic-aware Recommendation Model with Two-headed Autoencoder,TAAE)。该模型通过隐狄利克雷主题模型与BERT模型提取出用户评论的主题信息与语义信息,采用注意力机制与门控机制相结合的方式进行多模态特征对齐与融合,再利用多项式降噪自编码器进行用户评分预测。此外,为了缓解自编码器推荐模型中常见的流行度偏差问题,TAAE构建了一个负采样解码器,对推荐结果进一步优化。最后,在6组公开Amazon数据集上测试了TAAE模型的推荐性能,并对模型可能存在的变体及各解码器结构进行消融实验,实验结果表明,TAAE模型优于其他7个对比模型。

Abstract

Recommender system has become an indispensable part of many information service websites, such as e-commerce websites, online streaming medias and news distribution platforms. To alleviate the data sparsity and cold-start issues, we propose a two-headed autoencoder based topic-aware recommendation model for reviews to efficiently learn the representations of users and items. It firsty exploits Latent Dirichlet Allocation and BERT to learn topic and semantic information from users’ reviews, so as to achieve multi-modal information alignment and fusion with attention and gating mechanisms. To predict ratings for potential items via a polynomial denoising autoencoder, we employ a negative sampling decoder to solve the popularity bias in most autoencoder-based recommender systems. Experiments on four public datasets show that our proposed model outperforms seven baseline models.

关键词

双头自编码器 / 门控机制 / 协同过滤 / 评论感知 / 推荐系统

Key words

two-headed autoencoder / gating mechanism / collaborative filtering / review-aware / recommender system

引用本文

EndNote

Ris (Procite)

Bibtex

导出引用
刘树栋,李震,郝熙平,陈旭. 基于双头自编码器的评论主题感知推荐模型. 中文信息学报. 2024, 38(9): 146-166
LIU Shudong, LI Zhen, HAO Xiping, CHEN Xu. Two-headed Autoencoder Based Topic-aware Recommendation Model for Reviews. Journal of Chinese Information Processing. 2024, 38(9): 146-166

参考文献

[1] PANNIELLO U, TUZHILIN A, GORGOGLIO N M. Experimental comparison of pre-vs. post-filtering approaches in context-aware recommender systems [C]// Proceedings of the 3rd ACM Conference on Recommender Systems, 2009: 265-268.
[2] 孟祥武, 刘树栋, 张玉洁,等. 社会化推荐系统研究[J]. 软件学报, 2015, 26(6): 1356-1372.
[3] SHAN Y, HOENS T, JIAO J, et al. Deep crossing: Web-scale modeling without manually crafted combinatorial features [C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016: 255-262.
[4] CHENG H, KOC L, HARMSEN J, et al. Wide & deep learning for recommender systems [C]// Proceedings of the 1st Workshop on Deep Learning for Recommender Systems, 2016: 7-10.
[5] HE X, LIAO L, ZHANG H, et al. Neural collaborative filtering [C]// Proceedings of the 26th International Conference on World Wide Web, 2017: 173 -182.
[6] CANTADOR I, CREMONESI P. Tutorial on cross-domain recommender systems[C]// Proceedings of the 8th ACM Conference on Recommender systems, 2014: 401-402.
[7] TANG J, WU S, SUN J, et al. Cross-domain collaboration recommendation [C]// Proceedings of the 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2012: 1285-1293.
[8] MAN T, SHEN H, JIN X, et al. Cross-domain recommendation: An embedding and mapping approach [C]// Proceedings of the 26th International Joint Conference on Artificial Intelligence, 2017: 2464-2470.
[9] CAO J, LIN X, CONG X, et al. DisenCDR: Learning disentangled representations for cross-domain recommendation [C]// Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022: 267-277.
[10] WU S, SUN F, ZHANG W, et al. Graph neural networks in recommender systems: A survey[J]. ACM Computing Survey, 2023, 55(5): 1-97.
[11] YING R, HE R, CHEN K, et al. Graph convolutional neural networks for web-scale recommender systems[C]// Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018: 974-983.
[12] HE X, DENG K, WANG X, et al. LightGCN: Simplifying and powering graph convolution network for recommendation[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, 2020: 639-648.
[13] ZHANG M, WU S, YU X, et al. Dynamic graph neural networks for sequential recommendation[J]. IEEE Transactions on Knowledge and Data Engineering,2023,5: 4741 -4753.
[14] CHEN L, CHEN G, WANG F. Recommender systems based on user reviews: The state of the art[J]. User Model and User-adapted Interaction, 2015, 25: 99-154.
[15] GUANG L, LYU M, KING I. Ratings meet reviews, a combined approach to recommend [C]// Proceedings of the 8th ACM Conference on Recommender Systems, 2014: 105-112.
[16] CHENG Z, DING Y, ZHU L, et al. Aspect-aware latent factor model: Rating prediction with ratings and reviews [C]// Proceedings of the World Wide Web Conference, 2018: 639-648.
[17] JIN M, LUO X, ZHU H, et al. Combining deep learning and topic modeling for review understanding in context-aware recommendation [C]// Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics, 2018: 1605-1614.
[18] ZHANG G, LIU Y, JIN X. A survey of autoencoder-based recommender systems[J]. Frontiers of Computer Science, 2020, 14(2): 430-450.
[19] SEDHAIN S, MENON A K, SANNER S, et al. Autorec: Autoencoders meet collaborative filtering [C]// Proceedings of the 24th International Conference on World Wide Web, 2015: 111-112.
[20] LEE J. IDAE: Imputation-boosted denoising autoencoder for collaborative filtering [C]// Proceedings of the ACM on Conference on Information and Knowledge Management, 2017: 2143-2146.
[21] LIANG D, KRISHNAN R, HOFFMAN M, et al. Variational autoencoders for collaborative filtering [C]// Proceedings of the World Wide Web Conference, 2018: 689-698.
[22] YU X, ZHANG X, CAO Y, et al. VAEGAN: A collaborative filtering framework based on adversarial variational autoencoders [C]// Proceedings of the 68th International Joint Conference on Artificial Intelligence, 2019: 4206-4212.
[23] TRUONG Q, SALAH A, LAUW H. Bilateral variational autoencoder for collaborative filtering [C]// Proceedings of the 14th ACM International Conference on Web Search and Data Mining, 2021: 292-300.
[24] VANCURA V, ALVES R, KASALICKY P, et al. Scalable linear shallow autoencoder for collaborative filtering [C]// Proceedings of the 16th ACM Conference on Recommen der Systems, 2022: 604-609.
[25] MA C, KANG P, WU B, et al. Gated attentive-autoencoder for content-aware recommendation [C]// Proceedings of the 12th ACM International Conference on Web Search and Data Mining, 2019: 519-527.
[26] ZHOU J, CHENG Z, PEREZ F, et al. TAFA: Two-headed attention fused autoencoder for context-aware recommendations [C]// Proceedings of the 14th ACM Conference on Recommender Systems, 2020: 338-347.
[27] BATHLA G, AGGARWAL H, RANI R. AutoTrustRec: Recommender system with social trust and deep learning using autoencoder [J]. Multimedia Tools and Applications, 2020, 79: 20845-20860.
[28] XIE Z, LIU C, ZHANG Y, et al. Adversarial and contrastive variational autoencoder for sequential recommendation [C]// Proceedings of the Web Conference, 2021: 449-459.
[29] WANG Y, ZHANG H, LIU Z, et al. ContrastVAE: Contrastive variational autoEncoder for sequential recommendation [C]// Proceedings of the 31st ACM International Conference on Information & Knowledge Management, 2022: 2056-2066.
[30] LI L, XIAHOU J, LIN F, et al. DistVAE: Distributed variational autoencoder for sequential recommendation[J]. Knowledge-based Systems,2023,264: 110313.
[31] ZHOU F, WEN Z, ZHANG K, et al. Variational session-based recommendation using normalizing flows [C]// Proceedings of the World Wide Web Conference, 2019: 3476-3475.
[32] CHENG Z, DING Y, ZHU L, et al. Aspect-aware latent factor model: Rating prediction with ratings and reviews [C]// Proceedings of the World Wide Web Conference, 2018: 639-648.
[33] GUAN X, CHENG Z, HE X, et al. Attentive aspect modeling for review-aware recommendation[J]. ACM Transactions on Information Systems, 2019, 37(3): 28-36.
[34] LIU H, QIAO H, SHI X, et al. Aspect-aware asymmetric representation learning network for review-based recommendation [C]// Proceedings of International Joint Conference on Neural Networks, 2022: 1-8.
[35] SHUAI J, ZHANG K, WU L, et al. A Review-aware graph contrastive learning framework for recommendation[C]// Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022: 1283-1293.
[36] WANG K, ZHU Y, LIU H, et al. Learning aspect-aware high-order representations from ratings and reviews for recommendation[J]. ACM Transactions on Knowledge Discovery from Data, 2023, 17(1): 1-22.
[37] ZHANG Y. Incorporating phrase-level sentiment analysis on textual reviews for personalized recommendation[C]// Proceedings of the 8th ACM International Conference on Web Search and Data Mining, 2015: 435-440.
[38] HYUN D, PARK C, YANG M. Review sentiment-guided scalable deep recommender system [C]// Proceedings of the 41st International ACM SIGIR Conference on Research & Development in Information Retrieval, 2018: 965-968.
[39] HUANG C, JIANG W, WU J, et al. Personalized review recommendation based on users’ aspect sentiment [J]. ACM Transactions on Internet Technology, 2021,20 (4): 1-26.
[40] LIU P, ZHANG L, GULLA J. Multilingual review-aware deep recommender system via aspect-based sentiment analysis[J]. ACM Transactions on Information Systems, 2021, 39(2): 15-26.
[41] PARK S, CHAE D, BAE H, et al. Reinforcement learning over sentiment-augmented knowledge graphs towards accurate and explainable recommendation [C]// Proceedings of the 15th ACM International Conference on Web Search and Data Mining, 2022: 784-793.
[42] ZHOU J, WU G, MAI Z, et al. Noise contrastive estimation for autoencoding-based on-class collaborative filtering[J]. arXiv preprint arXiv: 2008.01246v2, 2020.
[43] XU Z, CHEN C, LUKASIEWICZ T, et al. Tag-aware personalized recommendation using a deep-semantic similarity model with negative sampling [C]// Proceedings of the 25th ACM International on Conference on Information and Knowledge Management, 2016: 1921-1924.
[44] CHEN C, MA W, ZHANG M, et al. Revisiting negative sampling vs. non-sampling in implicit recommendation[J]. ACM Transactions on Information Systems, 2023, 41(1): 1-25.
[45] SACHDEVA N, MANCO G, RITACCO E, et al. Sequential variational autoencoders for collaborative filtering [C]// Proceedings of the 12th ACM International Conference on Web Search and Data Mining, 2019: 600-608.
[46] SHENBIN I, ALEKSEEV A, TUTUBALINA E, et al. Recvae: A new variational autoencoder for top-n recommendations with implicit feedback [C]// Proceedings of the 13th International Conference on Web Search and Data Mining, 2020: 528-536.
[47] XIE Z, LIU C, ZHANG Y, et al. Adversarial and contrastive variational autoencoder for sequential recommendation [C]// Proceedings of the Web Conference, 2021: 449-459.

基金

国家自然科学基金(61602518,71872180);国家社会科学基金(21BXW076);高等学校学科创新引智基地(B21038);中南财经政法大学中央高校基本科研业务费专项资金(2722023EJ002)
PDF(3456 KB)

303

Accesses

0

Citation

Detail
段落导航
相关文章

  • 版权所有 © 《中文信息学报》编辑部 网站统计

    地址：北京市海淀区中关村南四街4号　邮编：100190　电话：010-62562916　E-mail：cips@iscas.ac.cn

    本系统由北京玛格泰克科技发展有限公司设计开发

/

〈 〉