PRoH: Dynamic Planning and Reasoning over Knowledge Hypergraphs for Retrieval-Augmented Generation

2025-10-14

AI Summary

Existing KH-based RAG methods suffer from three key limitations in multi-hop question answering: static retrieval strategies, non-adaptive execution, and shallow exploitation of structural semantics. To address these, we propose PRoH—a novel framework that pioneers the integration of dynamic programming and structured reasoning into knowledge hypergraph (KH) RAG. PRoH introduces a context-aware planning module for task-adaptive problem decomposition; constructs a directed acyclic graph (DAG) to support multi-granularity decomposition; and proposes an entity-weighted overlap-guided path retrieval algorithm that deeply fuses KH topological structure with semantic information. Empirically, PRoH achieves state-of-the-art performance across multiple benchmarks, improving F1 by 19.73% over HyperGraphRAG and enhancing generation quality by 8.41%. Moreover, it demonstrates strong robustness on long-distance multi-hop tasks, significantly advancing both reasoning depth and generation alignment.

Technology Category

• Knowledge Representation and Reasoning: Knowledge Engineering
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Search and Optimization: Metareasoning and Metaheuristics

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Knowledge Hypergraphs (KHs) have recently emerged as a knowledge representation for retrieval-augmented generation (RAG), offering a paradigm to model multi-entity relations into a structured form. However, existing KH-based RAG methods suffer from three major limitations: static retrieval planning, non-adaptive retrieval execution, and superficial use of KH structure and semantics, which constrain their ability to perform effective multi-hop question answering. To overcome these limitations, we propose PRoH, a dynamic Planning and Reasoning over Knowledge Hypergraphs framework. PRoH incorporates three core innovations: (i) a context-aware planning module that sketches the local KH neighborhood to guide structurally grounded reasoning plan generation; (ii) a structured question decomposition process that organizes subquestions as a dynamically evolving Directed Acyclic Graph (DAG) to enable adaptive, multi-trajectory exploration; and (iii) an Entity-Weighted Overlap (EWO)-guided reasoning path retrieval algorithm that prioritizes semantically coherent hyperedge traversals. Experiments across multiple domains demonstrate that PRoH achieves state-of-the-art performance, surpassing the prior SOTA model HyperGraphRAG by an average of 19.73% in F1 and 8.41% in Generation Evaluation (G-E) score, while maintaining strong robustness in long-range multi-hop reasoning tasks.

Problem

• Dynamic planning over knowledge hypergraphs for multi-hop question answering
• Overcoming static retrieval and non-adaptive execution limitations in RAG
• Addressing superficial use of knowledge structure and semantics

Innovation

• Context-aware planning module sketches local KH neighborhood
• Structured question decomposition uses dynamic DAG organization
• Entity-Weighted Overlap algorithm guides reasoning path retrieval