WWW 2026 | LLM×Graph论文总结【LLM4Graph & Graph4LLM】

时空探索之旅

发布于 2026-03-10 16:14:37

3060

WWW 2026将在2026年4月13日到17日于阿联酋迪拜（Dubai, United Arab Emirates）举行。

本文总结了2026 WWW上有关LLM Graph的相关论文，包含Research一个Track的论文（没有其它track），总计24篇，如有疏漏，欢迎补充。

笔者将LLM和Graph结合的工作分为两大类，一类是LLM4Graph，即LLM做图任务。其中细分了：文本属性图（Text-attributed Graph, TAG）,知识图谱（KG），图基础模型（GFM）。另外一类是利用Graph4LLM，即利用图这种格式来增强LLM的能力。

LLM4Graph1. A Graph Foundation Model for Unified Anomaly Detection2. RAG-GFM: Overcoming In-Memory Bottlenecks in Graph Foundation Models via Retrieval-Augmented Generation3. Disentangled Graph LLM for Molecule Graph Editing under Distribution Shifts4. Towards Graph Foundation Model: Node Feature Transfer Invariant Modeling on General Graphs5. Detecting Miscitation on the Scholarly Web through LLM-Augmented Text-Rich Graph Learning6. Text-attributed Graph Condensation via Text Selection and Attribute Matching7. Can LLMs Fool Graph Learning? Exploring Universal Adversarial Attacks on Text-Attributed Graphs8. UTAG: Leveraging LLM as a Unified Embedding Generator for Text-Attributed Graphs9. MixRAG : Mixture-of-Experts Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering10. Reasoning by Exploration: A Unified Approach to Retrieval and Generation over Graphs11. LLM-enhanced Federated Graph Learning with Geometry-aware Graph Projection and Shared Subspace Aggregation12. Node Role-Guided LLMs for Dynamic Graph Clustering13. A Unified Framework for Context-Aware and Relation-Aware Graph Retrieval-Augmented Generation14. A Unified Framework for Rule Learning: Integrating Commonsense Knowledge from LLMs with Structured Knowledge from Knowledge Graphs15. Towards Robust Detection of Chinese Toxic Variants via Dynamic Knowledge Graph–LLM Reasoning16. Towards Foundation Models for MMKG: Multi-Task Inductive Generalization via Task-Aware Routing17. Towards Open-World Retrieval-Augmented Generation on Knowledge Graph: A Multi-Agent Collaboration Framework18. RPO-RAG: Aligning Small LLMs with Relation-aware Preference Optimization for Knowledge Graph Question Answering19. ReaLM: Residual Quantization Bridges Knowledge Graph Embeddings and Large Language Models20. KG-BiLM: Knowledge Graph Embedding via Bidirectional Language Models21. VL-KGE: Vision-Language Models Meet Knowledge Graph EmbeddingGraph4LLM22. How Human Experts Educate Specialized LLMs: Filling Knowledge Gaps in KG-Augmented Generation through Hallucination Detection23. FraudShield: Knowledge Graph Empowered Defense for LLMs against Fraud Attacks24. MemoTime: Memory-Augmented Temporal Knowledge Graph Enhanced Large Language Model Reasoning

1 A Graph Foundation Model for Unified Anomaly Detection

关键词：异常检测，图基础模型

2 RAG-GFM: Overcoming In-Memory Bottlenecks in Graph Foundation Models via Retrieval-Augmented Generation

链接：https://arxiv.org/abs/2601.15124

作者：Haonan Yuan, Qingyun Sun, Jiacheng Tao, Xingcheng Fu and Jianxin Li

关键词：RAG，图基础模型，图提示学习

3 Disentangled Graph LLM for Molecule Graph Editing under Distribution Shifts

作者：Yang Yao, Xin Wang, Yuan Meng, Zeyang Zhang, Hong Mei and Wenwu Zhu

关键词：分子图，分布偏移

4 Towards Graph Foundation Model: Node Feature Transfer Invariant Modeling on General Graphs

作者：Jitao Zhao, Yi Wang, Yawen Li, Dongxiao He, Di Jin, Zhiyong Feng and Weixiong Zhang

关键词：图基础模型，迁移不变性

5 Detecting Miscitation on the Scholarly Web through LLM-Augmented Text-Rich Graph Learning

作者：Huidong Wu, Haojia Xiang, Jingtong Gao, Xiangyu Zhao, Dengsheng Wu and Jianping Li

关键词：错误学术引用，富文本图，LLM

6 Text-attributed Graph Condensation via Text Selection and Attribute Matching

作者：Haowei Han, Yuxiang Wang, Guojia Wan, Hao Wang, Shanshan Feng, Hao Huang, Jiawei Jiang and Xiao Yan

关键词：文本图压缩，属性匹配

7 Can LLMs Fool Graph Learning? Exploring Universal Adversarial Attacks on Text-Attributed Graphs

作者：Zihui Chen, Yuling Wang, Pengfei Jiao, Kai Wu, Xiao Wang, Xiang Ao and Dalin Zhang

关键词：文本图，对抗攻击，LLM

8 UTAG: Leveraging LLM as a Unified Embedding Generator for Text-Attributed Graphs

作者：Mingqian Ding, Jianjun Li, Zhiyuan Ma, Liwei Zhang and Wenqi Yang

关键词：文本图，嵌入表示，LLM

9 MixRAG : Mixture-of-Experts Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering

链接：https://arxiv.org/abs/2509.21391

作者：Lihui Liu, Jiayuan Ding, Subhabrata Mukherjee and Carl Yang

关键词：文本图问答，RAG

10 Reasoning by Exploration: A Unified Approach to Retrieval and Generation over Graphs

链接：https://arxiv.org/abs/2510.07484

作者：Haoyu Han, Kai Guo, Harry Shomer, Yu Wang, Yucheng Chu, Hang Li, Li Ma and Jiliang Tang

关键词：LLM，RAG，多跳问答

11 LLM-enhanced Federated Graph Learning with Geometry-aware Graph Projection and Shared Subspace Aggregation

作者：Pengyang Zhou, Zhihao Huang, Jiahe Xu, Wu Wen, Xiaolin Zheng, Chaochao Chen and Jianwei Yin

关键词：联邦图学习，LLM

12 Node Role-Guided LLMs for Dynamic Graph Clustering

作者：Dongyuan Li, Ying Zhang, Yaozu Wu and Renhe Jiang

关键词：动态图聚类，LLM

13 A Unified Framework for Context-Aware and Relation-Aware Graph Retrieval-Augmented Generation

作者：Haoyang Zhong, Yifei Sun, Antong Zhang, Chunping Wang, Lei Chen and Yang Yang

关键词：GraphRAG，上下文感知

14 A Unified Framework for Rule Learning: Integrating Commonsense Knowledge from LLMs with Structured Knowledge from Knowledge Graphs

作者：Qirui Hao, Kewei Cheng, Tongze Zhang, Hongyuan Liu, Junming Shao and Carl Yang

关键词：知识图谱，LLM

15 Towards Robust Detection of Chinese Toxic Variants via Dynamic Knowledge Graph–LLM Reasoning

作者：Shaochen Yang, Kefei Zhou and Wei Xu

关键词：动态知识图谱，LLM

16 Towards Foundation Models for MMKG: Multi-Task Inductive Generalization via Task-Aware Routing

作者：Shundong Yang, Jing Yang, Xiaowen Jiang, Xiaofen Wang, Laurence T. Yang, Yuan Gao, Xinfa Jiang, Jie Chen and Chaojun Zhang

关键词：MMKG，多任务，基础模型

17 Towards Open-World Retrieval-Augmented Generation on Knowledge Graph: A Multi-Agent Collaboration Framework

链接：https://arxiv.org/abs/2509.01238

作者：Jiasheng Xu, Mingda Li, Yongqiang Tang, Peijie Wang and Wensheng Zhang

关键词：RAG，LLM，Agent，知识图谱

18 RPO-RAG: Aligning Small LLMs with Relation-aware Preference Optimization for Knowledge Graph Question Answering

链接：https://arxiv.org/abs/2601.19225

作者：Kaehyun Um, Kyuhwan Yeom, Haerim Yang, Minyoung Choi, Hyeongjun Yang and Kyong-Ho Lee

关键词：知识图谱问答，RAG，LLM

19 ReaLM: Residual Quantization Bridges Knowledge Graph Embeddings and Large Language Models

链接：https://arxiv.org/abs/2510.09711

作者：Wenbin Guo, Xin Wang, Jiaoyan Chen, Lingbing Guo, Zhao Li and Zirui Chen

关键词：知识图谱补全，LLM

20 KG-BiLM: Knowledge Graph Embedding via Bidirectional Language Models

链接：https://arxiv.org/abs/2506.03576

作者：Zirui Chen, Xin Wang, Zhao Li, Wenbin Guo, Dongxiao He, Yanbing Li and Wushour Silamu

关键词：知识图谱嵌入，双向LLM

21 VL-KGE: Vision鈥揕anguage Models Meet Knowledge Graph Embedding

作者：Athanasios Efthymiou, Stevan Rudinac, Monika Kackovic, Nachoem Wijnberg and Marcel Worring

关键词：知识图谱嵌入，VLM

22 How Human Experts Educate Specialized LLMs: Filling Knowledge Gaps in KG-Augmented Generation through Hallucination Detection

作者：Chaofan Li, Lixing Chen, Junhua Tang, Yang Bai, Yutong Zhang, Zhi Zheng, Pan Zhou and Zhe Qu

关键词：幻觉检测，知识图谱增强生成

23 FraudShield: Knowledge Graph Empowered Defense for LLMs against Fraud Attacks

作者：Naen Xu, Jinghuai Zhang, Ping He, Chunyi Zhou, Jun Wang, Zhihui Fu, Tianyu Du, Zhaoxiang Wang and Shouling Ji

关键词：LLM欺诈检测，知识图谱增强

24 MemoTime: Memory-Augmented Temporal Knowledge Graph Enhanced Large Language Model Reasoning

链接：https://arxiv.org/abs/2510.13614

作者：ingyu Tan, Xiaoyang Wang, Qing Liu, Xiwei Xu, Xin Yuan, Liming Zhu and Wenjie Zhang

关键词：时序知识图谱增强推理，LLM，RAG