From Retrieval to Translation: Translating Query into Graph-level Clues for Retrieval-Augmented Generation

Retrieval-Augmented Generation (RAG) has recently been enhanced with tree or graph structures to match user intent for precise passage retrieval, which facilitates large language models (LLMs) in effectively mitigating hallucinations by leveraging external knowledge. However, we identify that existing structure-augmented RAG systems are experiencing (i) potential retrieval suspension and (ii) cumulative semantic drift, due to low-quality structures and semantic embeddings that often poorly capture textual details. Motivated by this, we propose a novel paradigm named KG-Translator, which is distinct from traditional matching-based paradigms and instead translates user queries into graph-level clues. Specifically, KG-Translator utilizes lightweight models to conduct named entity recognition (NER) and syntactic parsing on the corpus, constructing a reliable knowledge graph (ParseKG). On top of ParseKG, KG-Translator adopts constrained decoding strategies to faithfully translate clues, traces them to original passages, and employs a lightweight ranking model for precise passage retrieval. Extensive experiments on five datasets demonstrate that KG-Translator significantly outperforms baselines.

People like to use graph structures to store and search text knowledge, as they build many connections between knowledge. However, these connections also offer "support" for biased ideas to spread, since "no connections" and "wrong connections" greatly affect how we find knowledge. We use large language models as the "translator" between user searches and graphs. We keep graph designs simple, and mainly use large language models’ strong ability to explore graphs step by step. Surprisingly, this "translator" can find target knowledge precisely, even better than existing methods.