Elamur Research⌕
☰
home / papers / 2026 / Dynamic Graph Navigation via Triplet Chains for Structure-Aware Retrieval-Augmented Generation

Dynamic Graph Navigation via Triplet Chains for Structure-Aware Retrieval-Augmented Generation

Paper · 2026 · Feng Zhao, Yufei Wu, Xianggan Liu et al.

Summary

Standard retrieval-augmented generation (RAG) has two weaknesses. For easy questions, retrieval often adds irrelevant passages that can mislead the model. For hard multi-hop questions, the retrieved text is long, unstructured and scatters the key facts. Feng Zhao et al. propose Trig-Nav, which combines knowledge graphs with raw text. The pipeline has three stages:

  1. An LLM judges whether the retrieved context is sufficient. Simple questions are answered directly.
  2. Complex questions are split into sub-questions, and a targeted knowledge graph is built from triples extracted from the retrieved documents. Each sub-question is then turned into a chain of triples.
  3. Retrieval follows separate paths for factual, associative and causal knowledge. The retrieved triples are linked back to supporting text, and the answers to the sub-questions guide the final response.

Across four QA benchmarks, Trig-Nav beats single- and multi-source retrieval baselines, with about 6% average improvement over plain RAG. The work suggests that matching retrieval to question complexity and knowledge type works better than simply retrieving more passages.

Key points

  • Benchmarks and results: Trig-Nav was evaluated on NQ, TriviaQA, HotpotQA and 2WikiMultihopQA using Exact Match (EM), with roughly 6% average gain over traditional RAG.
  • Retriever robustness: Results stay stable across dense, BM25 and hybrid retrievers. EM ranges from 67.0% to 68.6% on TriviaQA and from 43.8% to 46.4% on HotpotQA.
  • Ablations: Removing question decomposition costs up to about 8%. The necessity-judgment module reduces per-question latency by skipping unneeded retrieval.
  • Effect of top-k: Larger top-k adds little or even hurts on single-hop datasets but helps on multi-hop ones. Between k=9 and k=21, performance rises and then falls, which the authors attribute to accumulating noise. Trends were confirmed with Qwen-7B.
  • Limitations: The method depends on the LLM's internal knowledge and its ability to follow prompts. It does not improve the model's inherent multi-hop reasoning.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Huazhong University of Science and TechnologyWuhan · China59 works · 8,310 citations

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksComputer Vision and Pattern Recognition · 3,486 worksInformation Systems · 1,730 works

🧭 Topics 1Research subjects and languages the work is about.Information retrieval & search · 552 works

Cite

Cite as Feng Zhao, Yufei Wu, Xianggan Liu and Ruilin Zhao. 2026. Dynamic Graph Navigation via Triplet Chains for Structure-Aware Retrieval-Augmented Generation. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.1360

@inproceedings{zhao2026dynamic,
  title = {{Dynamic Graph Navigation via Triplet Chains for Structure-Aware Retrieval-Augmented Generation}},
  author = {Feng Zhao, Yufei Wu, Xianggan Liu and Ruilin Zhao},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1360},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1360}
}

References

41 references, 23 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, and Hannaneh Hajishirzi. 2023. Self-rag: Learning to retrieve, generate, and critique through self-reflection. Preprint, arXiv:2310.11511.
  2. Arian Askari, Amin Abolghasemi, Gabriella Pasi, Wessel Kraaij, and Suzan Verberne. 2023. Injecting the bm25 score as text improves bert-based re-rankers. In European Conference on Information Retrieval, pages 66–83. Springer. find
  3. Sebastian Borgeaud, Arthur Mensch, Jordan Hoffmann, Trevor Cai, Eliza Rutherford, Katie Millican, George van den Driessche, Jean-Baptiste Lespiau, Bogdan Damoc, Aidan Clark, Diego de Las Casas, Aurelia Guy, Jacob Menick, Roman Ring, Tom Hennigan, Saffron Huang, Loren Maggiore, Chris Jones, Albin Cas… find
  4. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, and 1 others. 2020. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901.
  5. Yuzheng Cai, Zhenyue Guo, Yiwen Pei, Wanrui Bian, and Weiguo Zheng. 2025. Simgrag: Leveraging similar subgraphs for knowledge graphs driven retrieval-augmented generation. In Findings of the Association for Computational Linguistics: ACL 2025, pages 3139–3158. find
  6. Chia-Yuan Chang, Zhimeng Jiang, Vineeth Rakesh, Menghai Pan, Chin-Chia Michael Yeh, Guanchu Wang, Mingzhi Hu, Zhichao Xu, Yan Zheng, Mahashweta Das, and 1 others. 2025. Main-rag: Multi-agent filtering retrieval-augmented generation. In Proceedings of the 63rd Annual Meeting of the Association for Co…
  7. Zhanpeng Chen, Chengjin Xu, Yiyan Qi, and Jian Guo. 2024. Mllm is a strong reranker: Advancing multimodal retrieval-augmented generation via knowledge-enhanced reranking and noise-injected training. Preprint, arXiv:2407.21439. — [arXiv:2407.21439](https://arxiv.org/abs/2407.21439) find
  8. Luyu Gao, Xueguang Ma, Jimmy Lin, and Jamie Callan. 2022. Precise Zero-Shot Dense Retrieval without Relevance Labels. arXiv e-prints, arXiv:2212.10496.
  9. Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat, and Ming-Wei Chang. 2020. REALM: Retrieval-Augmented Language Model Pre-Training. arXiv e-prints, arXiv:2002.08909.
  10. Jingxuan Han, Zhendong Mao, Yi Liu, Yexuan Che, Zheren Fu, and Quan Wang. 2025. Fine-grained knowledge enhancement for retrieval-augmented generation. In Findings of the Association for Computational Linguistics: ACL 2025, pages 10031–10044. find
  11. Xanh Ho, Anh-Khoa Duong Nguyen, Saku Sugawara, and Akiko Aizawa. 2020. Constructing a multi-hop qa dataset for comprehensive evaluation of reasoning steps. arXiv preprint arXiv:2011.01060.
  12. Soyeong Jeong, Jinheon Baek, Sukmin Cho, Sung Ju Hwang, and Jong C. Park. 2024. Adaptive-rag: Learning to adapt retrieval-augmented large language models through question complexity. Preprint, arXiv:2403.14403.
  13. Zhengbao Jiang, Frank F. Xu, Luyu Gao, Zhiqing Sun, Qian Liu, Jane Dwivedi-Yu, Yiming Yang, Jamie Callan, and Graham Neubig. 2023. Active Retrieval Augmented Generation. arXiv e-prints, arXiv:2305.06983.
  14. Mandar Joshi, Eunsol Choi, Daniel S Weld, and Luke Zettlemoyer. 2017. Triviaqa: A large scale distantly supervised challenge dataset for reading comprehension. arXiv preprint arXiv:1705.03551.
  15. Saurav Kadavath, Tom Conerly, Amanda Askell, Tom Henighan, Dawn Drain, Ethan Perez, Nicholas Schiefer, Zac Hatfield-Dodds, Nova DasSarma, Eli Tran-Johnson, and 1 others. 2022. Language models (mostly) know what they know. arXiv preprint arXiv:2207.05221. find
  16. Urvashi Khandelwal, Omer Levy, Dan Jurafsky, Luke Zettlemoyer, and Mike Lewis. 2019. Generalization through Memorization: Nearest Neighbor Language Models. arXiv e-prints, arXiv:1911.00172. find
  17. Tom Kwiatkowski, Jennimaria Palomaki, Olivia Redfield, Michael Collins, Ankur Parikh, Chris Alberti, Danielle Epstein, Illia Polosukhin, Jacob Devlin, Kenton Lee, and 1 others. 2019. Natural questions: a benchmark for question answering research. Transactions of the Association for Computational Lin…
  18. Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, and 1 others. 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in neural information processing systems, 33:9459–…
  19. Xi Victoria Lin, Xilun Chen, Mingda Chen, Weijia Shi, Maria Lomeli, Rich James, Pedro Rodriguez, Jacob Kahn, Gergely Szilvasy, Mike Lewis, Luke Zettlemoyer, and Scott Yih. 2024. Ra-dit: Retrieval-augmented dual instruction tuning. Preprint, arXiv:2310.01352.
  20. Zihan Liu, Wei Ping, Rajarshi Roy, Peng Xu, Chankyu Lee, Mohammad Shoeybi, and Bryan Catanzaro. 2024. Chatqa: Surpassing gpt-4 on conversational qa and rag. Preprint, arXiv:2401.10225.
  21. Xinbei Ma, Yeyun Gong, Pengcheng He, Hai Zhao, and Nan Duan. 2023. Query rewriting for retrieval-augmented large language models. Preprint, arXiv:2305.14283. — [arXiv:2305.14283](https://arxiv.org/abs/2305.14283) find
  22. Joshua Maynez, Shashi Narayan, Bernd Bohnet, and Ryan McDonald. 2020. On faithfulness and factuality in abstractive summarization. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 1906–1919, Online. Association for Computational Linguistics.
  23. Shiyu Ni, Keping Bi, Jiafeng Guo, and Xueqi Cheng. 2024. When do llms need retrieval augmentation? mitigating llms’ overconfidence helps retrieval augmentation. Preprint, arXiv:2402.11457. find
  24. OpenAI, Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, Red Avila, Igor Babuschkin, Suchir Balaji, Valerie Balcom, Paul Baltescu, Haiming Bao, Mohammad Bavarian, Jeff Belgum, and 262 others.… — GPT-4 Technical Report
  25. Dongyu Ru, Lin Qiu, Xiangkun Hu, Tianhang Zhang, Peng Shi, Shuaichen Chang, Cheng Jiayang, Cunxiang Wang, Shichao Sun, Huanyu Li, Zizhao Zhang, Binjie Wang, Jiarong Jiang, Tong He, Zhiguo Wang, Pengfei Liu, Yue Zhang, and Zheng Zhang. 2024. Ragchecker: A fine-grained framework for diagnosing retriev… — [arXiv:2408.08067](https://arxiv.org/abs/2408.08067) find
  26. Zhihong Shao, Yeyun Gong, Yelong Shen, Minlie Huang, Nan Duan, and Weizhu Chen. 2023. Enhancing retrieval-augmented large language models with iterative retrieval-generation synergy. Preprint, arXiv:2305.15294.
  27. Freda Shi, Xinyun Chen, Kanishka Misra, Nathan Scales, David Dohan, Ed Chi, Nathanael Schärli, and Denny Zhou. 2023. Large Language Models Can Be Easily Distracted by Irrelevant Context. arXiv e-prints, arXiv:2302.00093.
  28. Weijia Shi, Sewon Min, Michihiro Yasunaga, Minjoon Seo, Richard James, Mike Lewis, Luke Zettlemoyer, and Wen-tau Yih. 2024. REPLUG: Retrieval-augmented black-box language models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Hum…
  29. Weihang Su, Yichen Tang, Qingyao Ai, Zhijing Wu, and Yiqun Liu. 2024. Dragin: dynamic retrieval augmented generation based on the information needs of large language models. arXiv preprint arXiv:2403.10081. — [arXiv:2403.10081](https://arxiv.org/abs/2403.10081) find
  30. Haoyu Wang, Ruirui Li, Haoming Jiang, Jinjin Tian, Zhengyang Wang, Chen Luo, Xianfeng Tang, Monica Xiao Cheng, Tuo Zhao, and Jing Gao. 2024. Blendfilter: Advancing retrieval-augmented large language models via query generation blending and knowledge filtering. In Proceedings of the 2024 Conference o… find
  31. Zhuocheng Zhang, Yang Feng, and Min Zhang. 2025. Levelrag: Enhancing retrieval-augmented generation with multi-hop logic planning over rewriting augmented searchers. Preprint, arXiv:2502.18139. — [arXiv:2502.18139](https://arxiv.org/abs/2502.18139) find
  32. Chunting Zhou, Graham Neubig, Jiatao Gu, Mona Diab, Paco Guzman, Luke Zettlemoyer, and Marjan Ghazvininejad. 2021. Detecting hallucinated content in conditional neural sequence generation. Preprint, arXiv:2011.02593.
  33. Yujia Zhou, Zheng Liu, Jiajie Jin, Jian-Yun Nie, and Zhicheng Dou. 2024. Metacognitive retrieval-augmented large language models. Preprint, arXiv:2402.11626. — [arXiv:2402.11626](https://arxiv.org/abs/2402.11626) find
  34. Xiangrong Zhu, Yuexiang Xie, Yi Liu, Yaliang Li, and Wei Hu. 2025. Knowledge graph-guided retrieval augmented generation. Preprint, arXiv:2502.06864. find
  35. Shijie Wang, Wenqi Fan, Yue Feng, Lin Shanru, Xinyu Ma, Shuaiqiang Wang, and Dawei Yin. 2025. Knowledge graph retrieval-augmented generation for llm-based recommendation. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 27152–…
  36. Yile Wang, Peng Li, Maosong Sun, and Yang Liu. 2023. Self-knowledge guided retrieval augmentation for large language models. Preprint, arXiv:2310.05002.
  37. Derong Xu, Xinhang Li, Ziheng Zhang, Zhenxi Lin, Zhihong Zhu, Zhi Zheng, Xian Wu, Xiangyu Zhao, Tong Xu, and Enhong Chen. 2025. Harnessing large language models for knowledge graph question answering via adaptive multi-aspect retrieval-augmentation. In Proceedings of the AAAI Conference on Artificia… find
  38. Zhilin Yang, Peng Qi, Saizheng Zhang, Yoshua Bengio, William Cohen, Ruslan Salakhutdinov, and Christopher D Manning. 2018. Hotpotqa: A dataset for diverse, explainable multi-hop question answering. In Proceedings of the 2018 conference on empirical methods in natural language processing, pages 2369–…
  39. Ori Yoran, Tomer Wolfson, Ori Ram, and Jonathan Berant. 2024. Making retrieval-augmented language models robust to irrelevant context. Preprint, arXiv:2310.01558.
  40. Yue Yu, Wei Ping, Zihan Liu, Boxin Wang, Jiaxuan You, Chao Zhang, Mohammad Shoeybi, and Bryan Catanzaro. 2024. Rankrag: Unifying context ranking with retrieval-augmented generation in llms. Preprint, arXiv:2407.02485. — [arXiv:2407.02485](https://arxiv.org/abs/2407.02485) find
  41. Yanzhao Zhang, Dingkun Long, Guangwei Xu, and Pengjun Xie. 2022. Hlatr: Enhance multi-stage text find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhao-2026-dynamic-graph-navigation-triplet-chains/