NaviRAG: Towards Active Knowledge Navigation for Retrieval-Augmented Generation

TL;DR

NaviRAG introduces an active, hierarchical knowledge navigation framework for retrieval-augmented generation, enhancing retrieval accuracy and answer quality in complex, long-document tasks.

Contribution

It presents a novel hierarchical knowledge structuring and active navigation approach that improves over traditional flat retrieval methods in RAG systems.

Findings

Improves retrieval recall and answer accuracy on long-document QA benchmarks.

Demonstrates the effectiveness of multi-granular evidence localization.

Ablation studies show dynamic retrieval planning enhances performance.

Abstract

Retrieval-augmented generation (RAG) typically relies on a flat retrieval paradigm that maps queries directly to static, isolated text segments. This approach struggles with more complex tasks that require the conditional retrieval and dynamic synthesis of information across different levels of granularity (e.g., from broad concepts to specific evidence). To bridge this gap, we introduce NaviRAG, a novel framework that shifts from passive segment retrieval to active knowledge navigation. NaviRAG first structures the knowledge documents into a hierarchical form, preserving semantic relationships from coarse-grained topics to fine-grained details. Leveraging this reorganized knowledge records, a large language model (LLM) agent actively navigates the records, iteratively identifying information gaps and retrieving relevant content from the most appropriate granularity level. Extensive experiments on long-document QA benchmarks show that NaviRAG consistently improves both retrieval recall and end-to-end answer performance over conventional RAG baselines. Ablation studies confirm performance gains stem from our method's capacity for multi-granular evidence localization and dynamic retrieval planning. We further discuss efficiency, applicable scenario, and future directions of our method, hoping to make RAG systems more intelligent and autonomous.