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

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.

Authors

• Jie Zhang University of Bath
• Bo Tang
• Wanzi Shao College of Computer and Data Science, Fuzhou University
• Wenqiang Wei MemTensor (Shanghai) Technology Co., Ltd.
• Jihao Zhao Renmin University of China
• Jianqing Zhu College of Engineering, Huaqiao University, Quanzhou 362021, China
• Zhiyu Li Shanghai Algorithm Innovation Research Institute
• Wen Xi China Haisum Engineering CO., LTD.
• Zehao Lin Institute for Advanced Algorithms Research, Shanghai, China
• Feiyu Xiong Institute for Advanced Algorithms Research, Shanghai, China
• Yanchao Tan College of Computer and Data Science, Fuzhou University, Fuzhou, China Fujian Provincial Key Laboratory of Network Computing and Intelligent Information Processing, Fuzhou University, Fuzhou, China

Keywords

No keywords are indexed for this paper.