大语言模型增强的时间注意力推荐系统

孙浩然; 王欣; 熊菲

doi:10.7652/xjtuxb202510021

您当前的位置：

首页 >

文章列表页 >

大语言模型增强的时间注意力推荐系统

更新时间：2025-09-28

- 大语言模型增强的时间注意力推荐系统
- Large Language Model-Augmented Time-Attention Recommender Systems
- 西安交通大学学报 2025年59卷第10期页码：221-230
- 作者机构：
- 作者简介：
  
  作者简介：孙浩然（1999—），男，博士生；
  熊菲（通信作者），男，教授，博士生导师。
- 基金信息：
  
  国家自然科学基金资助项目(62472024);中央高校基本科研业务费专项基金资助项目。(2024YJS203)
- DOI：10.7652/xjtuxb202510021
  中图分类号： TP391
- 收稿：2025-03-30，
  
  纸质出版：2025-10-10
- 稿件说明：
移动端阅览
孙浩然, 王欣, 熊菲. 大语言模型增强的时间注意力推荐系统[J]. 西安交通大学学报, 2025,59(10):221-230.

SUN Haoran, WANG Xin, XIONG Fei. Large Language Model-Augmented Time-Attention Recommender Systems[J]. Journal of Xi'an Jiaotong University, 2025, 59(10): 221-230.
孙浩然, 王欣, 熊菲. 大语言模型增强的时间注意力推荐系统[J]. 西安交通大学学报, 2025,59(10):221-230. DOI： 10.7652/xjtuxb202510021.

SUN Haoran, WANG Xin, XIONG Fei. Large Language Model-Augmented Time-Attention Recommender Systems[J]. Journal of Xi'an Jiaotong University, 2025, 59(10): 221-230. DOI： 10.7652/xjtuxb202510021.

摘要

为了解决传统推荐方法依赖于用户与物品的稀疏交互数据，其难以深入挖掘用户偏好背后的语义逻辑及其随时间演化的动态特征，以及大语言模型（LLM）在推荐系统中的直接应用受限于缺乏结构化交互建模与时间敏感性考虑的问题，提出了一种大语言模型增强的基于时间的推荐模型（LLATR）。LLATR旨在融合语义理解能力与用户兴趣的时间变化建模，以提升推荐精度与系统响应的个性化水平。模型设计了协同特征提取网络和时间特征建模网络，以及结合大语言模型生成的语义评分向量，通过对比学习机制，实现了多模态信息的统一建模，从而构建了具有动态适应能力的推荐框架。结果表明：在MovieLens-100K、Kaggle-Movie这2个数据集上，LLATR的均方根误差、平均绝对误差相对于现有主流模型提升了2%～5%。进一步分析表明，LLM能够补充协同特征之外的深层语义信息，并增强推荐系统对冷启动用户、稀疏数据和复杂行为背景的适应能力，以及有效建模用户的兴趣随时间的非线性演化趋势。

Abstract

To address the limitations of traditional recommendation methods that rely on sparse user-item interaction data—which struggle to deeply mine the semantic logic behind user preferences and their dynamic temporal evolution—as well as the constraints of directly applying large language models (LLMs) in recommendation systems due to their lack of structured interaction modeling and temporal sensitivity

a large language model-augmented time-aware recommendation (LLATR) algorithm is proposed. This algorithm aims to integrate semantic comprehension capabilities with temporal modeling of user interests to enhance recommendation accuracy and system personalization. The method designs a collaborative feature extraction network and a time feature modeling network

and combines the semantic scoring vectors generated by LLMs. Through a contrastive learning mechanism

it achieves unified modeling of multimodal information

thereby constructing a dynamically adaptive recommendation framework. Experiments are conducted on two datasets: MovieLens-100K andKaggle-Movie. The results demonstrate that LLATR improves root mean square error (RMSE) and mean absolute error (MAE) by 2%—5%compared to existing mainstream models. Further analysis reveals that LLMs can supplement deep semantic information beyond collaborative features

enhance the recommendation system's adaptability to cold-start users

sparse data

and complex behavioral contexts

and effectively model the nonlinear evolution trends of user interests over time.

关键词

Keywords

references

刘华玲 , 马俊 , 张国祥 . 基于深度学习的内容推荐算法研究综述 [J ] . 计算机工程 , 2021 , 47 ( 7 ): 1 - 12 .

LIU Hualing , MA Jun , ZHANG Guoxiang . Review of studies on deep learning-based content recommendation algorithms [J ] . Computer Engineering , 2021 , 47 ( 7 ): 1 - 12 .

王立才 , 孟祥武 , 张玉洁 . 上下文感知推荐系统 [J ] . 软件学报 , 2012 , 23 ( 1 ): 1 - 20 .

WANG Licai , MENG Xiangwu , ZHANG Yujie . Context-aware recommender systems [J ] . Journal of Software , 2012 , 23 ( 1 ): 1 - 20 .

陈海涵 , 吴国栋 , 李景霞 , 等 . 基于注意力机制的深度学习推荐研究进展 [J ] . 计算机工程与科学 , 2021 , 43 ( 2 ): 370 - 380 .

CHEN Haihan , WU Guodong , LI Jingxia , et al . Research advances on deep learning recommendation based on attention mechanism [J ] . Computer Engineering & Science , 2021 , 43 ( 2 ): 370 - 380 .

秦继伟 , 郑庆华 , 郑德立 , 等 . 结合评分和信任的协同推荐算法 [J ] . 西安交通大学学报 , 2013 , 47 ( 4 ): 100 - 104 .

QIN Jiwei , ZHENG Qinghua , ZHENG Deli , et al . A collaborative recommendation algorithm based on ratings and trust [J ] . Journal of Xi'an Jiaotong University , 2013 , 47 ( 4 ): 100 - 104 .

彭梓航 , 张全贵 , 金海波 , 等 . 基于时间块动态图神经网络的序列推荐方法 [J/OL ] . 计算机应用研究 .[ 2025-01-27 ] . https://doi.org/10.19734/j.issn.1001-3695.2024.12.0518 .

PENG Zihang , ZHANG Quangui , JIN Haibo , et al . Time-block-based dynamic graph neural network for sequential recommendation [J/OL ] . Application Research of Computers .[ 2025-01-27 ] . https://doi.org/10.19734/j.issn.1001-3695.2024.12.0518 .

GRAVES A . Supervised sequence labelling with recurrent neural networks [M ] . Berlin,Germany : Springer , 2012 .

吴文隆 , 尹海莲 , 王宁 , 等 . 大语言模型和知识图谱协同的跨域异质数据查询框架 [J ] . 计算机研究与发展 , 2025 , 62 ( 3 ): 605 - 619 .

WU Wenlong , YIN Hailian , WANG Ning , et al . A synergetic LLM-KG framework for cross-domain heterogeneous data query [J ] . Journal of Computer Research and Development , 2025 , 62 ( 3 ): 605 - 619 .

吕学强 , 万甜 , 马登豪 , 等 . LLM-PKE:一种集成大模型与多特征网络的专利关键词提取方法研究 [J/OL ] . 数据分析与知识发现 .[ 2025-02-19 ] . http://kns.cnki.net/kcms/detail/10.1478.G2.20250219.1348.006.html .

LV Xueqiang , WAN Tian , MA Denghao , et al . LLM-PKE:research on a patent keywords extraction method integrating large language models and multi-feature networks [J/OL ] . Data Analysis and Knowledge Discovery .[ 2025-02-19 ] . http://kns.cnki.net/kcms/detail/10.1478.G2.20250219.1348.006.html .

WANG Changdong , XI Wudong , HUANG Ling , et al . A BP neural network based recommender framework with attention mechanism [J ] . IEEE Transactions on Knowledge and Data Engineering , 2022 , 34 ( 7 ): 3029 - 3043 .

黄文 , 李振江 . 传统模式与大语言模型下推荐系统的比较研究 [J ] . 软件导刊 , 2025 , 24 ( 2 ): 204 - 210 .

HUANG Wen , LI Zhenjiang . Recommendation systems:a comparative study of traditional models and pre-trained model approaches [J ] . Software Guide , 2025 , 24 ( 2 ): 204 - 210 .

HUANG Chao , XU Huance , XU Yong , et al . Knowledge-aware coupled graph neural network for social recommendation [C ] // Proceedings of the AAAI Conference on Artificial Intelligence . Palo Alto,CA, USA : AAAI Press , 2021 : 4115 - 4122 .

KANG Wangcheng , MCAULEY J . Self-attentive sequential recommendation [C ] // 2018 IEEE International Conference on Data Mining . Piscataway,NJ,USA : IEEE , 2018 : 197 - 206 .

BAHDANAU D , CHO K , BENGIO Y . Neural machine translation by jointly learning to align and translate [EB/OL ] .[ 2025-02-16 ] . https://arxiv.org/abs/1409.0473 .

GLOROT X , BORDES A , BENGIO Y . Deep sparse rectifier neural networks [C ] // Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics . Chia Laguna Resort,Sardinia, Italy : PMLR , 2011 : 315 - 323 .

DENG Sinuo , WU Lifang , SHI Ge , et al . Learning to compose diversified prompts for image emotion classification [J ] . Computational Visual Media , 2024 , 10 ( 6 ): 1169 - 1183 .

LIN Zihan , TIAN Changxin , HOU Yupeng , et al . Improving graph collaborative filtering with neighborhood-enriched contrastive learning [C ] // Proceedings of the ACM Web Conference 2022 . New York,USA : ACM , 2022 : 2320 - 2329 .

史加荣 , 王丹 , 尚凡华 , 等 . 随机梯度下降算法研究进展 [J ] . 自动化学报 , 2021 , 47 ( 9 ): 2103 - 2119 .

SHI Jiarong , WANG Dan , SHANG Fanhua , et al . Research advances on stochastic gradient descent algorithms [J ] . Acta Automatica Sinica , 2021 , 47 ( 9 ): 2103 - 2119 .

HARPER F M , KONSTAN J A . The MovieLens datasets:history and context [J ] . ACM Transactions on Interactive Intelligent Systems , 2015 , 5 ( 4 ): 19 .

LUO Sichun , ZHANG Xinyi , XIAO Yuanzhang , et al . HySAGE:a hybrid static and adaptive graph embedding network for context-drifting recommendations [C ] // Proceedings of the 31st ACM International Conference on Information & Knowledge Management . New York,USA : ACM , 2022 : 1389 - 1398 .

XIONG Fei , SUN Haoran , LUO Guixun , et al . Graph attention network with high-order neighbor information propagation for social recommendation [C ] // Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence (IJCAI-24) . California,USA : IJCAI , 2024 : 2478 - 2486 .

LIU Qiang , YU Feng , WU Shu , et al . A convolutional click prediction model [C ] // Proceedings of the 24th ACM International on Conference on Information and Knowledge Management . New York,USA : ACM , 2015 : 1743 - 1746 .

GUO Huifeng , TANG Ruiming , YE Yunming , et al . DeepFM:a factorization-machine based neural network for CTR prediction [C ] // Proceedings of the 26th International Joint Conference on Artificial Intelligence . Palo Alto,CA,USA : AAAI Press , 2017 : 1725 - 1731 .

HE Xiangnan , LIAO Lizi , ZHANG Hanwang , et al . Neural collaborative filtering [C ] // Proceedings of the 26th International Conference on World Wide Web . Republic and Canton of Geneva,CHE : International World Wide Web Conferences Steering Committee , 2017 : 173 - 182 .

WANG Ruoxi , FU Bin , FU Gang , et al . Deep &cross network for ad click predictions [C ] // Proceedings of theADKDD'17 . New York,USA : ACM , 2017 : 12 .

XIAO Jun , YE Hao , HE Xiangnan , et al . Attentional factorization machines:learning the weight of feature interactions via attention networks [C ] // Proceedings of the 26th International Joint Conference on Artificial Intelligence . Palo Alto, CA, USA : AAAI Press , 2017 : 3119 - 3125 .

LIAN Jianxun , ZHOU Xiaohuan , ZHANG Fuzheng , et al . xDeepFM:combining explicit and implicit feature interactions for recommender systems [C ] // Proceedings of the 24thACM SIGKDD International Conference on Knowledge Discovery & Data Mining . New York,USA : ACM , 2018 : 1754 - 1763 .

SONG Weiping , SHI Chence , XIAO Zhiping , et al . AutoInt: automatic feature interaction learning via self-attentive neural networks [C ] // Proceedings of the 28th ACM International Conference on Information and Knowledge Management . New York,USA : ACM , 2019 : 1161 - 1170 .

LUO Sichun , WANG Jiansheng , ZHOU Aojun , et al . Large language models augmented rating prediction in recommender system [C ] // ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing . Piscataway, NJ, USA : IEEE , 2024 : 7960 - 7964 .

TOUVRON H , LAVRIL T , IZACARD G , et al . Llama:open and efficient foundation language models [EB/OL ] .[ 2025-02-10 ] . https://arxiv.org/abs/2302.13971 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向高海拔空冷发电机防晕性能演化规律的电热耦合研究

水蒸气掺混对氨裂解气旋流燃烧与排放特性的影响研究

采用机器学习的大规模并行程序I/O性能建模与预测方法

混合电解槽阵列协同的风光耦合制氢系统双层多目标优化

超临界二氧化碳向心透平轮背间隙内风阻损失及流动特性研究