摘要 现有序列推荐方法未能充分挖掘项目属性语义信息且存在语义空间迁移不匹配的问题,导致对长尾物品推荐能力不足.为此,文中提出融合大模型语义增强信息和协同信息融合的序列推荐方法（Large Language Model Enhancement and Collaborative Information Fusion for Sequential Recommendation, LLM-CFSR）.首先,通过属性级数据增强与对比微调技术,利用大语言模型生成细粒度语义嵌入,捕捉长尾物品的深层语义关联.然后,设计双视图融合机制,分别从语义视图与协同视图两方面对用户偏好进行联合建模.最后,引入交叉注意力机制,实现嵌入层、序列层与预测层的多层次信息融合,促进语义信息与协同信号的深度交互.在Yelp、Amazon Fashion、Amazon Beauty数据集上的实验表明,LLM-CFSR对于整体推荐性能和长尾物品推荐性能都有所提升.
Abstract：Existing sequential recommendation methods fail to fully explore item attribute semantic information and suffer from a semantic space migration mismatch. These limitations result in inadequate recommendation capabilities of the existing methods for long-tail items. To address this issue, a method for sequential recommendation with large language model enhancement and collaborative information fusion（LLM-CFSR） is proposed in this paper. First, fine-grained semantic embeddings are generated with a large language model through attribute-level data augmentation and contrastive fine-tuning techniques to capture the deep semantic associations of long-tail items. Then, a dual-view modeling framework is designed to jointly model user preferences from both semantic and collaborative views. Finally, to promote deep interaction between semantic information and collaborative signals, a cross-attention mechanism is introduced to achieve multi-level information fusion across embedding layers, sequence layers, and prediction layers. Experimental results on Yelp, Amazon Fashion and Amazon Beauty datasets demonstrate that LLM-CFSR improves the overall recommendation performance and the long-tail item recommendation performance.
查龙宝, 黄琪, 王明文, 周骏祥, 罗文兵. 基于大语言模型增强和协同信息融合的序列推荐方法[J]. 模式识别与人工智能, 2025, 38(10): 925-937.
ZHA Longbao, HUANG Qi, WANG Mingwen, ZHOU Junxiang, LUO Wenbing. Sequential Recommendation with Large Language Model Enhancement and Collaborative Information Fusion. Pattern Recognition and Artificial Intelligence, 2025, 38(10): 925-937.
[1] MCAULEY J, TARGETT C, SHI Q F, et al. Image-Based Reco-mmendations on Styles and Substitutes//Proc of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2015: 43-52. [2] YIN J W, LIU C H, WANG W Q, et al. Learning Transferrable Parameters for Long-Tailed Sequential User Behavior Modeling//Proc of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York, USA: ACM, 2020: 359-367. [3] LIU Z W, FAN Z W, WANG Y, et al. Augmenting Sequential Re-commendation with Pseudo-Prior Items via Reversely Pre-training Transformer//Proc of the 44th International ACM SIGIR Confe-rence on Research and Development in Information Retrieval. New York, USA: ACM, 2021: 1608-1612. [4] 卡祖铭,赵鹏,张波,等.面向大语言模型的推荐系统综述.计算机科学, 2024, 51(S2): 11-21. (KA J M, ZHAO P, ZHANG B, et al. Survey of Recommendation Systems for Large Language Models. Computer Science, 2024, 51(S2): 11-21.) [5] TOUVRON H, LAVRIL T, IZACARD G, et al. LLaMA: Open and Efficient Foundation Language Models[C/OL].[2025-08-16]. https://arxiv.org/pdf/2302.13971. [6] HARTE J, ZORGDRAGER W, LOURIDAS P, et al. Leveraging Large Language Models for Sequential Recommendation//Proc of the 17th ACM Conference on Recommender Systems. New York, USA: ACM, 2023: 1096-1102. [7] BAO K Q, ZHANG J Z, WANG W J, et al. A Bi-step Grounding Paradigm for Large Language Models in Recommendation Systems. ACM Transactions on Recommender Systems, 2025, 3(4). DOI: 10.1145/3716393. [8] HU J, XIA W W, ZHANG X L, et al. Enhancing Sequential Re-commendation via LLM-Based Semantic Embedding Learning//Proc of the ACM Web Conference. New York, USA: ACM, 2024: 103-111. [9] LIU D G, XIAN S X, LIN X L, et al. A Practice-Friendly Two-Stage LLM-Enhanced Paradigm in Sequential Recommendation[C/OL].[2025-08-16]. https://arxiv.org/abs/2406.00333. [10] HIDASI B, KARATZOGLOU A.Recurrent Neural Networks with Top-k Gains for Session-Based Recommendations//Proc of the 27th ACM International Conference on Information and Knowledge Management. New York, USA: ACM, 2018: 843-852. [11] TANG J X, WANG K.Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding//Proc of the 11th ACM International Conference on Web Search and Data Mining. New York, USA: ACM, 2018: 565-573. [12] KANG W C, MCAULEY J.Self-Attentive Sequential Recommendation//Proc of the IEEE International Conference on Data Mi-ning. Washington, USA: IEEE, 2018: 197-206. [13] SUN F, LIU J, WU J, et al. BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer//Proc of the 28th ACM International Conference on Information and Knowledge Management. New York, USA: ACM, 2019: 1441-1450. [14] SCHNABEL T, SWAMINATHAN A, SINGH A, et al. Recommenda-tions as Treatments: Debiasing Learning and Evaluation//Proc of the 33rd International Conference on Machine Learning. San Diego, USA: PMLR, 2016: 1670-1679. [15] ZHOU K, WANG H, ZHAO W X, et al. S3-Rec: Self-Supervised Learning for Sequential Recommendation with Mutual Information Maximization//Proc of the 29th ACM International Conference on Information and Knowledge Management. New York, USA: ACM, 2020: 1893-1902. [16] YAO T S, YI X Y, CHENG D Z, et al. Self-Supervised Learning for Large-Scale Item Recommendations//Proc of the 30th ACM International Conference on Information and Knowledge Management. New York, USA: ACM, 2021: 4321-4330. [17] LIAO J Y, LI S H, YANG Z Y, et al. LLaRA: Large Language-Recommendation Assistant//Proc of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2024: 1785-1795. [18] 张凯涵,冯晨娇,姚凯旋,等.基于对比学习和语义增强的多模态推荐算法.模式识别与人工智能, 2024, 37(6): 479-490. (ZHANG K H, FENG C J, YAO K X, et al. Multimodal Reco-mmendation Algorithm Based on Contrastive Learning and Semantic Enhancement. Pattern Recognition and Artificial Intelligence, 2024, 37(6): 479-490.) [19] 吴国栋,秦辉,胡全兴,等.大语言模型及其个性化推荐研究.智能系统学报, 2024, 19(6): 1351-1365. (WU G D, QIN H, HU Q X, et al. Research on Large Language Models and Personalized Recommendation. CAAI Transactions on Intelligent Systems, 2024, 19(6): 1351-1365.) [20] DU H W, YUAN H H, ZHAO P P, et al. Ensemble Modeling with Contrastive Knowledge Distillation for Sequential Recommendation//Proc of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2023: 58-67. [21] BAO K Q, ZHANG J Z, ZHANG Y, et al. TALLRec: An Effective and Efficient Tuning Framework to Align Large Language Mo-del with Recommendation//Proc of the 17th ACM Conference on Recommender Systems. New York, USA: ACM, 2023: 1007-1014. [22] LIU Q D, WU X, WANG W Y, et al. LLMEmb: Large Language Model Can Be a Good Embedding Generator for Sequential Reco-mmendation. Proceedings of the AAAI Conference on Artificial Intelligence, 2025, 39(11): 12183-12191. [23] XIE X, SUN F, LIU Z Y, et al. Contrastive Learning for Sequential Recommendation//Proc of the IEEE 38th International Conference on Data Engineering. Washington, USA: IEEE, 2022: 1259-1273. [24] CHANG J X, GAO C, ZHENG Y, et al. Sequential Recommendation with Graph Neural Networks//Proc of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2021: 378-387. [25] JANG S, LEE H, CHO H, et al. CITIES: Contextual Inference of Tail-Item Embeddings for Sequential Recommendation//Proc of the IEEE International Conference on Data Mining. Washington, USA: IEEE, 2020: 202-211. [26] KIM K, HYUN D, YUN S, et al. MELT: Mutual Enhancement of Long-Tailed User and Item for Sequential Recommendation//Proc of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2023: 68-77. [27] LIU Q D, WU X, WANG Y J, et al. Large Language Models Enhancement for Long-Tailed Sequential Recommendation//Proc of the 38th International Conference on Neural Information Processing Systems. Cambridge, USA： MIT Press, 2024: 26701-26727.
2025, Vol. 38 
