Reason-Align-Respond: Aligning LLM Reasoning with Knowledge Graphs for KGQA

TL;DR

This paper introduces Reason-Align-Respond (RAR), a framework that combines large language model reasoning with knowledge graphs to improve question answering accuracy and interpretability.

Contribution

RAR is a novel probabilistic framework that aligns LLM reasoning with knowledge graph paths, enhancing KGQA performance and interpretability.

Findings

Achieves state-of-the-art Hit@1 scores of 93.3% on WebQSP and 91.0% on CWQ.

Generates high-quality, interpretable reasoning chains aligned with KG paths.

Exhibits strong zero-shot generalization and computational efficiency.

Abstract

LLMs have demonstrated remarkable capabilities in complex reasoning tasks, yet they often suffer from hallucinations and lack reliable factual grounding. Meanwhile, knowledge graphs (KGs) provide structured factual knowledge but lack the flexible reasoning abilities of LLMs. In this paper, we present Reason-Align-Respond (RAR), a novel framework that systematically integrates LLM reasoning with knowledge graphs for KGQA. Our approach consists of three key components: a Reasoner that generates human-like reasoning chains, an Aligner that maps these chains to valid KG paths, and a Responser that synthesizes the final answer. We formulate this process as a probabilistic model and optimize it using the Expectation-Maximization algorithm, which iteratively refines the reasoning chains and knowledge paths. Extensive experiments on multiple benchmarks demonstrate the effectiveness of RAR, achieving state-of-the-art performance with Hit@1 scores of 93.3% and 91.0% on WebQSP and CWQ respectively. Human evaluation confirms that RAR generates high-quality, interpretable reasoning chains well-aligned with KG paths. Furthermore, RAR exhibits strong zero-shot generalization capabilities and maintains computational efficiency during inference.