arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~

期刊&会议

ACM SIGIR Conference on Research and Development in Information Retrieval · 会议 · Information Retrieval

共收录 1420
2604.19186 2026-04-22 cs.LG cs.AI

Inductive Subgraphs as Shortcuts: Causal Disentanglement for Heterophilic Graph Learning

诱导子图作为捷径:异质图学习中的因果解缠

Xiangmeng Wang, Qian Li, Haiyang Xia, Hao Miao, Qing Li, Guandong Xu

机构 * The Hong Kong Polytechnic University(香港理工大学) Curtin University(Curtin大学) University of Macau(澳门大学) The Education University of Hong Kong(香港教育大学)

AI总结 本文提出因果解缠图神经网络CD-GNN,通过阻断非因果路径提升异质图节点分类的鲁棒性和准确性。

Comments SIGIR 2026

Journal ref SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.18146 2026-04-22 cs.IR cs.AI cs.CL

Modular Representation Compression: Adapting LLMs for Efficient and Effective Recommendations

模块化表示压缩:适应LLM以实现高效且有效的推荐

Yunjia Xi, Menghui Zhu, Jianghao Lin, Bo Chen, Ruiming Tang, Yong Yu, Weinan Zhang

机构 * Shanghai Jiao Tong University(上海交通大学) Huawei Noah's Ark Lab(华为诺亚实验室) Antai College of Economics and Management, Shanghai Jiao Tong University(上海交通大学安泰经济管理学院)

AI总结 本文提出MARC方法,通过模块化调整和任务解耦,解决LLM表示压缩中的MRA问题,提升推荐系统效率与效果。

Comments SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2401.16167 2026-04-22 cs.HC cs.CL

"You tell me": A Dataset of GPT-4-Based Behaviour Change Support Conversations

你告诉我:一个基于GPT-4的行为改变支持对话数据集

Selina Meyer, David Elsweiler

机构 * Regensburg University(莱茵河畔大学)

AI总结 本文提出一个基于GPT-4的对话数据集,用于研究用户行为对LLM生成文本的影响,包含用户互动、语言分析和反馈数据,为行为改变系统设计提供实证支持。

Comments Preprint as accepted at the 2024 ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR '24)

Journal ref In Proceedings of the 2024 Conference on Human Information Interaction and Retrieval (CHIIR '24). Association for Computing Machinery, New York, NY, USA, 411-416

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.18351 2026-04-21 cs.IR cs.LG

Balanced Co-Clustering of Users and Items for Embedding Table Compression in Recommender Systems

用户和物品的平衡协同聚类用于推荐系统中的嵌入表压缩

Runhao Jiang, Renchi Yang, Donghao Wu

机构 * Hong Kong Baptist University(香港 Baptist 大学) The Chinese University of Hong Kong(香港中文大学)

AI总结 本文提出BACO框架,通过协同信号对用户和物品进行聚类,减少嵌入表参数,提升推荐系统效率,实验显示参数减少75%且召回率下降仅1.85%。

Comments 14 pages, The technical report for the paper titled "Balanced Co-Clustering of Users and Items for Embedding Table Compression in Recommender Systems" in SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.18313 2026-04-21 cs.CV

Denoise and Align: Diffusion-Driven Foreground Knowledge Prompting for Open-Vocabulary Temporal Action Detection

去噪与对齐:基于扩散的前景知识提示用于开放词汇时序动作检测

Sa Zhu, Wanqian Zhang, Lin Wang, Jinchao Zhang, Cong Wang, Bo Li

机构 * Institute of Information Engineering, Chinese Academy of Sciences School of Cyber Security, University of Chinese Academy of Sciences State Key Laboratory of Cyberspace Security Defense Beijing China Institute of Information Engineering, Chinese Academy of Sciences Beijing China Hangzhou Dianzi University Hangzhou China Institute of Information Engineering, Chinese Academy of Sciences\ Key Laboratory of Cyberspace Security Defense Beijing China Engineering, Zhejiang University Hangzhou China Institute of Information Engineering, Chinese Academy of Sciences State Key Laboratory of Cyberspace Security Defense Beijing China Institute of Information Engineering, Chinese Academy of Sciences School of Cyber Security, University of Chinese Academy of Sciences State Key Laboratory of Cyberspace Security Defense Institute of Information Engineering, Chinese Academy of Sciences Hangzhou Dianzi University Institute of Information Engineering, Chinese Academy of Sciences\ Key Laboratory of Cyberspace Security Defense Engineering, Zhejiang University Institute of Information Engineering, Chinese Academy of Sciences State Key Laboratory of Cyberspace Security Defense

AI总结 本文提出DFAlign框架,通过扩散去噪生成前景知识,解决开放词汇时序动作检测中语义不平衡问题,提升动作相关片段的判别性。

Comments Accepted by SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.10947 2026-04-21 cs.IR

Multi-Faceted Continual Knowledge Graph Embedding for Semantic-Aware Link Prediction

多维持续知识图谱嵌入用于语义感知的链接预测

Jing Qi, Yuxiang Wang, Zhiyuan Yu, Xiaoliang Xu, Yuanshi Zheng, Tianxing Wu

AI总结 本文提出MF-CKGE框架,通过分离时间旧新知识和语义解耦提升空间效率,改进持续链接预测性能,实验显示在八个数据集上MRR和Hits@10分别提升1.7%(2.7%)和1.4%(3.8%)

Comments 11 pages, accepted by SIGIR 2026(full paper)

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.18731 2026-04-21 cs.CL cs.AI

One Adapts to Any: Meta Reward Modeling for Personalized LLM Alignment

一个适应任何:面向个性化大语言模型对齐的元奖励建模

Hongru Cai, Yongqi Li, Tiezheng Yu, Fengbin Zhu, Wenjie Wang, Fuli Feng, Wenjie Li

机构 * The Hong Kong Polytechnic University(香港理工大学) Huawei Technologies Ltd.(华为技术有限公司) National University of Singapore(新加坡国立大学) University of Science and Technology of China(中国科学技术大学)

AI总结 本文提出元奖励建模(MRM)方法,通过元学习框架优化奖励模型初始化,提升个性化对齐的效率与鲁棒性,实验表明其在少样本场景下表现优异。

Comments Accepted by SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.17405 2026-04-21 cs.AI

STRIDE: Strategic Iterative Decision-Making for Retrieval-Augmented Multi-Hop Question Answering

STRIDE: 用于检索增强多跳问答的策略迭代决策

Wei Chen, Lili Zhao, Zhi Zheng, HuiJun Hou, Tong Xu

机构 * University of Science Technology of China \& State Key Laboratory of Cognitive Intelligence Hefei Anhui China Technology of China \& State Key Laboratory of Cognitive Intelligence

AI总结 STRIDE通过分离战略规划、动态控制和 grounded 执行,解决多跳问答中实体早 Commit 和推理依赖缺失的问题,提升推理鲁棒性与准确性。

Comments Accepted by SIGIR 2026 Full Paper. The code repository is available at https://github.com/fanshu6hao/STRIDE

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.07825 2026-04-21 cs.IR cs.AI

Filling the Gaps: Selective Knowledge Augmentation for LLM Recommenders

填补空白:面向LLM推荐系统的选择性知识增强

Jaehyun Lee, Sanghwan Jang, SeongKu Kang, Hwanjo Yu

机构 * Pohang University of Science and Technology(釜山科学技术大学) Korea University(韩国大学)

AI总结 本文提出KnowSA_CKP方法,通过选择性注入知识来缓解LLM推荐中的知识差距问题,提升推荐准确性和上下文效率。

Comments Accepted to SIGIR 2026 full papers track

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.22913 2026-04-21 cs.IR cs.LG

SIGMA: A Semantic-Grounded Instruction-Driven Generative Multi-Task Recommender at AliExpress

SIGMA:阿里国际站的语义引导指令驱动生成多任务推荐系统

Yang Yu, Lei Kou, Huaikuan Yi, Bin Chen, Yayu Cao, Lei Shen, Chao Zhang, Bing Wang, Xiaoyi Zeng

机构 * Alibaba International Digital Commercial Group(阿里巴巴国际数字商业集团)

AI总结 SIGMA通过语义引导和指令驱动方法,解决传统推荐系统在多任务和实时业务需求中的局限,提升推荐准确性和多样性。

Comments Accepted by SIGIR 2026 Industry Track. 5 pages, 3 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.05336 2026-04-21 cs.CL cs.AI

WeatherArchive-Bench: Benchmarking Retrieval-Augmented Reasoning for Historical Weather Archives

WeatherArchive-Bench: 基于历史天气档案的检索增强推理基准测试

Yongan Yu, Xianda Du, Qingchen Hu, Jiahao Liang, Jingwei Ni, Dan Qiang, Kaiyu Huang, Grant McKenzie, Renee Sieber, Fengran Mo

机构 * McGill University(麦吉尔大学) University of Waterloo(滑铁卢大学) ETH Zurich(苏黎世联邦理工学院) Beijing Jiaotong University(北京交通大学)

AI总结 本文提出WeatherArchive-Bench,用于评估检索增强生成系统在处理历史天气档案中的能力,揭示密集检索器在历史术语上的不足及LLM对社会脆弱性与韧性概念的误判问题。

Comments accepted to the Resource Track of SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.16821 2026-04-21 cs.LG

R&F-Inventory: A Large-Scale Dataset for Monotonic Inventory Estimation in Reach and Frequency Advertising

R&F-Inventory:用于可达性与频率广告中单调库存估计的大型数据集

Yunshan Peng, Ji Wu, Wentao Bai, Yunke Bai, Jinan Pang, Wenzheng Shu, Yanxiang Zeng, Xialong Liu, Peng Jiang

机构 * Kuaishou Technology(快手科技)

AI总结 本文提出并发布了R&F合同库存估计数据集,用于研究结构约束学习、单调回归和R&F合同规划等问题,通过提供预算-性能曲线的完整数据支持系统研究。

Comments Accepted by SIGIR 2026; 7 pages

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.13743 2026-04-21 cs.CL cs.IR

LTRR: Learning To Rank Retrievers for LLMs

基于LLM的检索排名学习检索器

To Eun Kim, Fernando Diaz

机构 * Carnegie Mellon University(卡内基梅隆大学)

AI总结 本文提出LTRR框架,通过动态选择检索器提升RAG性能,实验表明基于路由的RAG在多种基准上优于单一检索器,尤其在使用AC目标和成对排名时效果显著。

Comments SIGIR 2026; SIGIR 2025 LiveRAG Spotlight; Code: https://github.com/kimdanny/Starlight-LiveRAG

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.07930 2026-04-20 cs.IR

Unified Supervision for Walmart's Sponsored Search Retrieval via Joint Semantic Relevance and Behavioral Engagement Modeling

通过联合语义相关性和行为参与建模实现沃尔玛赞助搜索检索的统一监督

Shasvat Desai, Md Omar Faruk Rokon, Jhalak Nilesh Acharya, Isha Shah, Hong Yao, Utkarsh Porwal, Kuang-chih Lee

AI总结 本文提出一种基于语义相关性和行为参与的统一监督框架,通过结合语义标签、检索先验分数和用户参与信号,提升沃尔玛电子商务赞助搜索检索性能。

Comments Accepted to SIGIR 2026, Industry Track

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.15882 2026-04-20 cs.IR cs.CL

JFinTEB: Japanese Financial Text Embedding Benchmark

JFinTEB:日本金融文本嵌入基准

Masahiro Suzuki, Hiroki Sakaji

机构 * Amova Asset Management Co., Ltd.(Amova资产管理部门有限公司) Hokkaido University(北海道大学)

AI总结 本文提出JFinTEB,首个针对日本金融文本嵌入评估的基准,涵盖检索与分类任务,评估多种嵌入模型,为日本金融文本挖掘社区提供标准化评估协议。

Comments 5 pages. Accepted at SIGIR 2026 Resource Track

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.19339 2026-04-20 cs.IR cs.AI cs.CL

Spectral Tempering for Embedding Compression in Dense Passage Retrieval

光谱温控用于密集路径检索中的嵌入压缩

Yongkang Li, Panagiotis Eustratiadis, Evangelos Kanoulas

机构 * University of Amsterdam(阿姆斯特丹大学)

AI总结 本文提出Spectral Tempering方法,通过分析语料库的光谱特性,自适应调整光谱缩放参数γ,实现高效的嵌入压缩,无需标注数据或验证搜索。

Comments This paper has been accepted as a short paper at SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.15190 2026-04-17 cs.AI cs.CL

Meituan Merchant Business Diagnosis via Policy-Guided Dual-Process User Simulation

美团商户业务诊断 via 政策引导的双过程用户模拟

Ziyang Chen, Renbing Chen, Daowei Li, Jinzhi Liao, Jiashen Sun, Ke Zeng, Xiang Zhao

机构 * Independent Researcher(独立研究者)

AI总结 本文提出Policy-Guided Hybrid Simulation框架,通过双过程融合提升商户策略评估的准确性,实验显示其在美团101家商户中误差降低45.8%。

Comments 5 pages, 3 figures, 2 tables, accepted at SIGIR 2026 Industry Track

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.14878 2026-04-17 cs.IR cs.AI

GenRec: A Preference-Oriented Generative Framework for Large-Scale Recommendation

GenRec:一种面向偏好的大规模推荐生成框架

Yanyan Zou, Junbo Qi, Lunsong Huang, Yu Li, Kewei Xu, Jiabao Gao, Binglei Zhao, Xuanhua Yang, Sulong Xu, Shengjie Li

机构 * Waseda University(早稻田大学)

AI总结 GenRec通过单解码器架构解决大规模推荐中生成检索的三个挑战,提出Page-wise NTP任务提升梯度信号,采用异构线性Token合并压缩输入,结合GRPO-SR强化学习方法提升用户满意度,实现点击量和交易量分别提升9.5%和8.7%。

Comments SIGIR 2026 Camera-Ready version

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.14839 2026-04-17 cs.IR

Well Begun is Half Done: Training-Free and Model-Agnostic Semantically Guaranteed User Representation Initialization for Multimodal Recommendation

万事开头难:免训练 且模型无关的语义保证用户表示初始化方法用于多模态推荐

Jinfeng Xu, Zheyu Chen, Shuo Yang, Jinze Li, Hewei Wang, Jianheng Tang, Wei Wang, Xiping Hu, Edith C. H. Ngai

AI总结 本文提出SG-URInit方法,通过整合用户交互物品的模态特征和全局聚类特征,实现免训练且模型无关的用户表示初始化,有效提升多模态推荐性能并缓解物品冷启动问题。

Comments Accepted by SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.14256 2026-04-17 cs.IR cs.AI

Evaluation of Agents under Simulated AI Marketplace Dynamics

对模拟AI市场动态下代理的评估

To Eun Kim, Alireza Salemi, Hamed Zamani, Fernando Diaz

机构 * Carnegie Mellon University(卡内基梅隆大学) University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校)

AI总结 本文提出Marketplace Evaluation框架,通过模拟市场动态评估信息访问系统,补充传统准确度指标,探讨市场留存和份额等指标。

Comments SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.14223 2026-04-17 cs.IR cs.AI

TRACE: A Conversational Framework for Sustainable Tourism Recommendation with Agentic Counterfactual Explanations

TRACE:一种用于可持续旅游推荐的对话框架,配备代理反事实解释

Ashmi Banerjee, Adithi Satish, Wolfgang Wörndl, Yashar Deldjoo

机构 * Technical University of Munich(慕尼黑技术大学) Polytechnic University of Bari(巴里理工学院)

AI总结 TRACE通过多代理架构促进可持续旅游,利用反事实解释和LLM生成问题,提升用户环保意识,实验证明其在推荐质量与交互响应上的有效性。

Journal ref Proceedings of the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '26), July 20--24, 2026, Melbourne, VIC, Australia

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.09982 2026-04-17 cs.IR cs.CL cs.LG

Reproduction Beyond Benchmarks: ConstBERT and ColBERT-v2 Across Backends and Query Distributions

超越基准的再现:ConstBERT和ColBERT-v2在不同后端和查询分布上的表现

Utshab Kumar Ghosh, Ashish David, Shubham Chatterjee

机构 * Missouri University of Science and Technology(密苏里科技大学)

AI总结 研究评估了ConstBERT和ColBERT-v2在不同后端和查询分布下的表现,发现ConstBERT在MS-MARCO上再现性良好,但在长叙事查询上性能显著下降,揭示了多向量检索架构的限制。

Comments 10 pages, 9 tables. Accepted to the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.21242 2026-04-17 cs.IR

Bi-Level Optimization for Generative Recommendation: Bridging Tokenization and Generation

生成推荐的双层优化:连接分词与生成

Yimeng Bai, Chang Liu, Yang Zhang, Dingxian Wang, Frank Yang, Andrew Rabinovich, Wenge Rong, Fuli Feng

AI总结 本文提出BLOGER框架,通过双层优化统一分词与推荐器,解决两者间依赖关系。实验表明其在多个真实数据集上优于现有方法,且计算效率高。

Comments Accepted by SIGIR'26

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13665 2026-04-16 cs.IR

RecNextEval: A Reference Implementation for Temporal Next-Batch Recommendation Evaluation

RecNextEval:用于时间下一批次推荐评估的参考实现

Tze-Kean Ng, Joshua Teng-Khing Khoo, Aixin Sun

AI总结 RecNextEval通过时间窗口数据分割确保模型在全局时间线上评估,减少数据泄露,提升推荐系统评估的准确性和可复现性。

Comments Accepted to SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13389 2026-04-16 cs.IR

RoTE: Coarse-to-Fine Multi-Level Rotary Time Embedding for Sequential Recommendation

RoTE: 从粗到细的多级旋转时间嵌入用于序列推荐

Haolin Zhang, Longtao Xiao, Guohao Cai, Ruixuan Li, Xiu Li

AI总结 RoTE通过多级时间嵌入模型捕捉用户时间跨度信息,提升序列推荐模型对长期和短期兴趣演变的捕捉能力,在三个基准测试中实现NDCG@5最高20.11%的提升。

Comments Accepted by SIGIR'26

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13273 2026-04-16 cs.IR

Mitigating Collaborative Semantic ID Staleness in Generative Retrieval

缓解生成检索中协同语义ID过时问题

Vladimir Baikalov, Iskander Bagautdinov, Sergey Muravyov

AI总结 本文提出轻量级、模型无关的SID对齐更新方法,解决时间漂移导致的SID过时问题,提升检索性能并降低训练计算量。

Comments Accepted at SIGIR 2026. This version corresponds to the accepted manuscript

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.00570 2026-04-16 cs.IR

SPRINT: Scalable and Predictive Intent Refinement for LLM-Enhanced Session-based Recommendation

SPRINT: 一种可扩展且可预测的意图细化方法用于基于会话的推荐系统

Gyuseok Lee, Wonbin Kweon, Zhenrui Yue, Yaokun Liu, Yifan Liu, Susik Yoon, Dong Wang, SeongKu Kang

AI总结 本文提出SPRINT框架,通过全局意图池约束LLM生成可靠意图,并在推理时使用轻量级意图预测器提升可扩展性,实验证明其在推荐效果和可解释性上优于现有方法。

Comments SIGIR'26

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.12990 2026-04-15 cs.IR

Sparse Contrastive Learning for Content-Based Cold Item Recommendation

稀疏对比学习用于基于内容的冷启动推荐

Gregor Meehan, Johan Pauwels

AI总结 本文提出基于内容的冷启动推荐方法,通过稀疏对比学习提升推荐准确性,优于传统方法。

Comments Accepted at SIGIR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.12201 2026-04-15 cs.IR

AdversarialCoT: Single-Document Retrieval Poisoning for LLM Reasoning

AdversarialCoT:单文档检索污染用于LLM推理

Hongru Song, Yu-An Liu, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Yixing Fan, Xueqi Cheng

AI总结 研究针对RAG系统中的知识库污染攻击,提出AdversarialCoT方法,通过单文档污染影响LLM推理,揭示其安全风险并提供改进方案。

Comments 6 pages,accepted by SIGIR 2026 short paper

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.11317 2026-04-15 cs.IR

Next Interest Flow: A Generative Pre-training Paradigm for Recommender Systems by Modeling All-domain Movelines

下一项兴趣流:通过建模所有领域运动线的生成预训练范式来构建推荐系统

Chen Gao, Zixin Zhao, Lv Shao, Tong Liu

AI总结 本文提出Next Interest Flow模型,通过建模用户兴趣的连续演化轨迹,解决传统方法在全局分布和结构演变建模上的不足,通过引入运动约束和双向对齐策略提升推荐效果。

Journal ref Proceedings of the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '26), July 20--24, 2026, Melbourne, VIC, Australia

详情

展开后加载摘要…

URL PDF HTML 收藏