ROG: Retrieval-Augmented LLM Reasoning for Complex First-Order Queries over Knowledge Graphs

TL;DR

ROG is a retrieval-augmented framework that enhances large language models' ability to reason over incomplete knowledge graphs for complex first-order logic queries by decomposing queries and grounding reasoning in relevant neighborhood evidence.

Contribution

It introduces ROG, a novel retrieval-augmented approach that decomposes complex FOL queries and grounds reasoning in neighborhood evidence, improving robustness and accuracy.

Findings

ROG outperforms embedding-based baselines on KG reasoning benchmarks.

Significant improvements on high-complexity queries.

Enhanced handling of negation-heavy queries.

Abstract

Answering first-order logic (FOL) queries over incomplete knowledge graphs (KGs) is difficult, especially for complex query structures that compose projection, intersection, union, and negation. We propose ROG, a retrieval-augmented framework that combines query-aware neighborhood retrieval with large language model (LLM) chain-of-thought reasoning. ROG decomposes a multi-operator query into a sequence of single-operator sub-queries and grounds each step in compact, query-relevant neighborhood evidence. Intermediate answer sets are cached and reused across steps, improving consistency on deep reasoning chains. This design reduces compounding errors and yields more robust inference on complex and negation-heavy queries. Overall, ROG provides a practical alternative to embedding-based logical reasoning by replacing learned operators with retrieval-grounded, step-wise inference. Experiments on standard KG reasoning benchmarks show consistent gains over strong embedding-based baselines, with the largest improvements on high-complexity and negation-heavy query types.