Knowledge Graph-Guided Retrieval Augmented Generation

By Xiangrong Zhu, Yuexiang Xie, Yi Liu, Ya-Liang Li, Wei Hu. 2025

Subfields: Retrieval-Augmented Generation, Knowledge Graphs & Representation

Citations in the corpus

• A Matryoshka Hierarchical RAG for Efficient Multi-Hop Question Answering (2026, unknown): “Another way of proceeding exploits the LLM as a reasoning engine to navigate the graph step by step, as in Think-on-Graph (ToG) [32] and Graph-CoT [14], or to extract task-relevant substructures, as in SG-RAG [29] and KG 2 RAG [45].”
• SemRAG: Semantic Knowledge-Augmented RAG for Improved Question-Answering (2025, unknown): “During generation, retrieved graph elements augment prompts, enabling the model to generate more contextually accurate and informed responses [20].”