CCSFR:基于协同-内容语义融合的评论增强推荐模型

doi:10.11896/jsjkx.250800027

摘要/Abstract

摘要： 基于评论的混合推荐系统将用户评论整合到协同过滤方法中来挖掘用户兴趣偏好的语义方面。然而,现有方法往往忽略了评论语义与用户交互之间的语义鸿沟,且未能充分挖掘两者的内在相似性。此外,当前混合推荐系统通常将协同信号与评论语义深度耦合,导致难以实现细粒度的推荐建模。对此,提出了基于协同-内容语义融合的评论增强推荐模型(CCSFR)。该模型利用协同信号与评论语义之间的相似性学习更全面的用户兴趣与物品特征,并增强用户和物品的嵌入表示。为了充分利用两者的相似性,设计了混合内容表示学习模块解耦协同过滤与基于内容的推荐,分别将用户交互与评论语义建模为交互图和方面图;随后利用图卷积神经网络捕获协同信号与语义特征,从而学习用户和物品在协同模式和内容模式中的嵌入表示。在此基础上,设计了协同语义融合策略,通过整合异构特征获得统一且高质量的表示。该策略通过语义对齐,将协同模式与内容模式统一在共享语义空间,以减小协同模式和内容模式之间的语义差异。此外,采用图对比学习,通过最大化协同模式与内容模式的互信息来捕获两者之间的相似性,促进两者互补融合,提升表示质量。在3个数据集上进行的实验结果表明,CCSFR在nDCG指标上相比最佳基线平均提升了8.5%和5.8%;在召回率指标上平均提升了8.2%和4.6%。消融实验进一步验证了该方法各个模块的有效性。实验结果充分表明,协同语义融合策略能够有效整合协同信号与评论语义,生成更具判别性的个性化表示,从而提升推荐性能。

关键词: 推荐系统, 对比学习, 图神经网络, 语义对齐, 异构融合

Abstract: Review-based hybrid recommendation systems explore the semantic aspects of user preferences by integrating user reviews into collaborative filtering methods.However,these methods ignore the semantic gap between the semantic reviews and user interactions,and thus insufficiently exploit the inherent similarities among them.On the other hand,most existing hybrid re-commendations deeply couple collaborative signals with semantic reviews,which fails to model recommendations in a fine-grained manner.To overcome these limitations,this paper proposes a collaborative-content semantic fusion for review-enhanced recommendation(CCSFR) model.The model leverages the similarity between collaborative signals and review semantics to learn more comprehensive user interests and item features,and to enhance the embedded representations of users and items.To fully leverage the similarities between collaborative signals and review semantics,a hybrid content representation learning module is designed to decouple collaborative filtering and content-based recommendation,modelling user interactions and review semantics as interaction graphs and aspect graphs,respectively.Subsequently,a graph convolutional neural network is utilized to capture both collaborative signals and semantic features to learn embedded representations of users and items in both collaborative and content modes.Building on this,a collaborative semantic fusion strategy is employed to obtain unified and high-quality representations by integrating heterogeneous features.This strategy unifies the collaborative and content modes in a shared semantic space through semantic alignment,thereby reducing the semantic gap between them.Furthermore,graph contrastive learning is adopted to capture the similarity between the two modes by optimizing the mutual information between both modes and promoting complementary integration to enhance the quality of the representation.Experiments on three datasets show that CCSFR improves average 8.5% and 5.8% on nDCG compared to the optimal baseline,and 8.2% and 4.6% on recall metrics.The results of these experiments fully demonstrate that the collaborative semantic fusion strategy can effectively integrate collaborative signals and review semantics to generate more discriminative personalized representations,thereby improving recommendation performance.

Key words: Recommendation systems, Contrastive learning, Graph neural network, Semantic alignment, Heterogeneous fusion

中图分类号:

TP391

周昊斌, 鲁云浩, 秦隽, 焦新涛, 曾碧卿. CCSFR:基于协同-内容语义融合的评论增强推荐模型[J]. 计算机科学, 2026, 53(8): 276-284. https://doi.org/10.11896/jsjkx.250800027

ZHOU Haobin, LU Yunhao, QIN Jun, JIAO Xintao, ZENG Biqing. CCSFR:Collaborative-Content Semantic Fusion for Review-enhanced Recommendation[J]. Computer Science, 2026, 53(8): 276-284. https://doi.org/10.11896/jsjkx.250800027

参考文献

[1] SHI X,ZHANG Y,PUJAHARI A,et al.When Latent Features Meet Side Information:A Preference Relation Based Graph Neural Network for Collaborative Filtering[J].Expert Systems with Applications,2025,260:125423.
[2] WANG C,ZHU H,ZHU C,et al.SetRank:A Setwise Bayesian Approach for Collaborative Ranking in Recommender System[J].ACM Transactions on Information Systems,2024,42(2):1-32.
[3] HE Z Y,LIN J Q,WANG C D,et al.Interaction-Knowledge Se-mantic Alignment for Recommendation[J].Neural Networks,2025,181:106755.
[4] XIONG Y,LIU Y,QIAN Y,et al.Review-based Recommenda-tion Under Preference Uncertainty:An Asymmetric Deep Learning Framework[J].European Journal of Operational Research,2024,316(3):1044-1057.
[5] HONG Y,YE W.Review-Enhanced Sequential Recommendation with Self-Attention and Graph Collaborative Features[C]//2023 IEEE International Conference on Data Mining Workshops(ICDMW).IEEE,2023:1493-1499.
[6] ZHAN Z,XU B.Analyzing Review Sentiments and ProductImages by Parallel Deep Nets for Personalized Recommendation[J].Information Processing & Management,2023,60(1):103166.
[7] LIU H,CHEN Y,LI P,et al.Enhancing Review-based User Representation on Learned Social Graph for Recommendation[J].Knowledge-Based Systems,2023,266:110438.
[8] LIU H,WANG W,PENG Q,et al.Toward Comprehensive User and Item Representations via Three-tier Attention Network[J].ACM Transactions on Information Systems,2021,39(3):1-22.
[9] WANG X,OUNIS I,MACDONALD C.Leveraging ReviewProperties for Effective Recommendation[C]//Proceedings of the Web Conference 2021.ACM,2021:2209-2219.
[10] YANG W,HUO T,LIU Z,et al.Review-based Multi-intention Contrastive Learning for Recommendation[C]//Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval(SIGIR 2023).New York:ACM,2023:2339-2343.
[11] PEINELT N,NGUYEN D,LIAKATA M.tBERT:Topic Mo-dels and BERT Joining Forces for Semantic Similarity Detection[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.2020:7047-7055.
[12] GHEEWALA S,XU S,YEOM S.In-depth Survey:Deep Lear-ning in Recommender Systems－Exploring Prediction and Ran-king Models,Datasets,Feature Analysis,and Emerging Trends[J].Neural Computing and Applications,2025,37(17):10875-10947.
[13] WEI Y,XU Y,ZHU L,et al.Multi-level Cross-modal Contrastive Learning for Review-aware Recommendation[J].Expert Systems with Applications,2024,247:123341.
[14] WANG K,ZHU Y,ZANG T,et al.Multi-aspect Graph Contrastive Learning for Review-enhanced Recommendation[J].ACM Transactions on Information Systems,2024,42(2):1-29.
[15] ZHANG C,XUE S,LI J,et al.Multi-aspect Enhanced GraphNeural Networks for Recommendation[J].Neural Networks,2023,157:90-102.
[16] ZHENG L,NOROOZI V,YU P S.Joint Deep Modeling ofUsers and Items Using Reviews for Recommendation[C]//Proceedings of the Tenth ACM International Conference on Web Search and Data Mining.ACM,2017:425-434.
[17] CHEN C,ZHANG M,LIU Y,et al.Neural Attentional Rating Regression with Review-level Explanations[C]//Proceedings of the 2018 World Wide Web Conference on World Wide Web－WWW ’18.ACM,2018:1583-1592.
[18] LIU Y,YANG S,ZHANG Y,et al.Learning Hierarchical Review Graph Representations for Recommendation[J].IEEE Transactions on Knowledge and Data Engineering,2023,35(1):658-671.
[19] ANELLI V W,YASHAR D,TOMMASO D N,et al.Reshaping Graph Recommendation with Edge Graph Collaborative Filtering and Customer Reviews[C]//Proceedings of the Workshop on Deep Learning for Search and Recommendation Colocated with the 31st International Conference on Information and Knowledge Management.CEUR-WS,2022:1-10.
[20] ZHAO Y,XU Q,ZOU Y,et al.Modeling User Reviews through Bayesian Graph Attention Networks for Recommendation[J].ACM Transactions on Information Systems,2023,41(3):1-29.
[21] LIU S,MENG Z,MACDONALD C,et al.Graph Neural Pre-training for Recommendation with Side Information[J].ACM Transactions on Information Systems,2023,41(3):1-28.
[22] YU J,XIA X,CHEN T,et al.XSimGCL:Towards ExtremelySimple Graph Contrastive Learning for Recommendation[J].IEEE Transactions on Knowledge and Data Engineering,2024,36(2):913-926.
[23] GAO C,ZHENG Y,LI N,et al.A Survey of Graph Neural Networks for Recommender Systems:Challenges,Methods,and Directions[J].ACM Transactions on Recommender Systems,2023,1(1):1-51.
[24] WANG X,HE X,WANG M,et al.Neural Graph Collaborative Filtering[C]//Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval.ACM,2019:165-174.
[25] HE X,DENG K,WANG X,et al.LightGCN:Simplifying andPowering Graph Convolution Network for Recommendation[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval.ACM,2020:639-648.
[26] JING M,ZHU Y,ZANG T,et al.Contrastive Self-supervised Learning in Recommender Systems:A Survey[J].ACM Transactions on Information Systems,2024,42(2):1-39.
[27] WU J,WANG X,FENG F,et al.Self-supervised Graph Lear-ning for Recommendation[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval.ACM,2021:726-735.
[28] YU J,YIN H,XIA X,et al.Are Graph Augmentations Necessary?:Simple Graph Contrastive Learning for Recommendation[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval.ACM,2022:1294-1303.
[29] LIN Z,TIAN C,HOU Y,et al.Improving Graph Collaborative Filtering with Neighborhood-enriched Contrastive Learning[C]//Proceedings of the ACM Web Conference 2022.ACM,2022:2320-2329.
[30] WANG X,JIN H,ZHANG A,et al.Disentangled Graph Colla-borative Filtering[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval.ACM,2020:1001-1010.
[31] TAHA K,YOO P D,YEUN C,et al.Empirical and Experimental Perspectives on Big Data in Recommendation Systems:A Comprehensive Survey[J].Big Data Mining and Analytics,2024,7(3):964-1014.
[32] YANG H,ZHANG C,LI K.PyABSA:A Modularized Framework for Reproducible Aspect-based Sentiment Analysis[C]//Proceedings of the 32nd ACM International Conference on Information and Knowledge Management.ACM,2023:5117-5122.
[33] REIMERS N,GUREVYCH I.Sentence-BERT:Sentence Em-beddings using Siamese BERT-networks[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing(EMNLP-IJCNLP).Association for Computational Linguistics,2019.
[34] YANG G,LI J,LI M,et al.Bridging Smoothness and Approximation:Theoretical Insights into Over-smoothing in Graph Neural Networks[J].arXiv:2407.01281,2024.
[35] CHEN L,WU L,HONG R,et al.Revisiting Graph Based Collaborative Filtering:A Linear Residual Graph Convolutional Network Approach[J].Proceedings of the AAAI Conference on Artificial Intelligence,2020,34(1):27-34.
[36] PATEL R,THAKKAR P,UKANI V.CNNRec:Convolutional Neural Network based recommender systems－A Survey[J].Engineering Applications of Artificial Intelligence,2024,133:108062.
[37] TANG H,LIU J,ZHAO M,et al.Progressive Layered Extraction(PLE):A Novel Multi-Task Learning(MTL) Model for Personalized Recommendations[C]//Fourteenth ACM Confe-rence on Recommender Systems.ACM,2020:269-278.
[38] WANG T,ISOLA P.Understanding Contrastive Representation Learning Through Alignment and Uniformity on the Hypersphere[C]//Proceedings of the 37th International Conference on Machine Learning.PMLR,2020:9929-9939.
[39] VAN DEN O,LI Y,VINYALS O.Representation Learningwith Contrastive Predictive Coding[J].arXiv:1807.03748,2018.
[40] RENDLE S,FREUDENTHALER C,GANTNER Z,et al.BPR:Bayesian Personalized Ranking from Implicit Feedback[C]//Proceedings of the Twenty-fifth Conference on Uncertainty in Artificial Intelligence.AUAI Press,2009:452-461.
[41] SHUAI J,ZHANG K,WU L,et al.A Review-aware GraphContrastive Learning Framework for Recommendation[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval.ACM,2022:1283-1293.
[42] HASAN E,RAHMAN M,DING C,et al.Review-based Recommender Systems:A Survey of Approaches,Challenges and Future Perspectives[J].ACM Computing Surveys,2025,58(1):1-41.
[43] ZHAO W X,CHEN J,WANG P,et al.Revisiting AlternativeExperimental Settings for Evaluating Top-N Item Recommendation Algorithms[C]//Proceedings of the 29th ACM International Conference on Information & Knowledge Management.ACM,2020:2329-2332.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed