EUVIMEDEUVIMED entdecken
Crossref

TAdaRAG: Task Adaptive Retrieval-Augmented Generation via On-the-Fly Knowledge Graph Construction

Jie Zhang, Bo Tang, Wanzi Shao, Wenqiang Wei, Jihao Zhao, Jianqing Zhu, Zhiyu Li, Wen Xi, Zehao Lin, Feiyu Xiong, Yanchao Tan

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (41) · 34728-34736

DOI: 10.1609/aaai.v40i41.40774 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Jie Zhang und 10 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Retrieval-Augmented Generation (RAG) improves large language models by retrieving external knowledge, often truncated into smaller chunks due to the input context window, which leads to information loss, resulting in response hallucinations and broken reasoning chains. Moreover, traditional RAG retrieves unstructured knowledge, introducing irrelevant details that hinder accurate reasoning. To address these issues, we propose TAdaRAG, a novel RAG framework for on-the-fly task-adaptive knowledge graph construction from external sources. Specifically, we design an intent-driven routing mechanism to a domain-specific extraction template, followed by supervised fine-tuning and a reinforcement learning-based implicit extraction mechanism, ensuring concise, coherent, and non-redundant knowledge integration. Evaluations on six public benchmarks and a real-world business benchmark (NowNewsQA) across three backbone models demonstrate that TAdaRAG outperforms existing methods across diverse domains and long-text tasks, highlighting its strong generalization and practical effectiveness.

Zitieren

APA

Zhang, J., Tang, B., Shao, W., Wei, W., Zhao, J., Zhu, J., Li, Z., Xi, W., Lin, Z., Xiong, F., & Tan, Y. (2026). TAdaRAG: Task Adaptive Retrieval-Augmented Generation via On-the-Fly Knowledge Graph Construction. Proceedings of the AAAI Conference on Artificial Intelligence, 40(41), 34728-34736. https://doi.org/10.1609/aaai.v40i41.40774

Vancouver

Zhang J, Tang B, Shao W, Wei W, Zhao J, Zhu J, et al. TAdaRAG: Task Adaptive Retrieval-Augmented Generation via On-the-Fly Knowledge Graph Construction. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(41):34728-34736. doi:10.1609/aaai.v40i41.40774

BibTeX
@article{Zhang2026,
  author = {Zhang, Jie and Tang, Bo and Shao, Wanzi and Wei, Wenqiang and Zhao, Jihao and Zhu, Jianqing and Li, Zhiyu and Xi, Wen and Lin, Zehao and Xiong, Feiyu and Tan, Yanchao},
  title = {TAdaRAG: Task Adaptive Retrieval-Augmented Generation via On-the-Fly Knowledge Graph Construction},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {41},
  pages = {34728--34736},
  doi = {10.1609/aaai.v40i41.40774}
}
Kostenlos starten →