UniRel: Relation-Centric Knowledge Graph Question Answering with RL-Tuned LLM Reasoning

TL;DR

UniRel is a novel framework for relation-centric KGQA that uses RL-tuned LLMs to identify informative, compact subgraphs representing entity relations, improving answer quality and generalization.

Contribution

This work introduces UniRel, a modular approach combining a subgraph retriever with RL-tuned LLMs for relation-centric KGQA, addressing the challenge of selecting meaningful subgraphs.

Findings

UniRel outperforms prompting baselines in connectivity and reward metrics.

The framework generalizes well to unseen entities and relations.

UniRel achieves competitive or improved performance in entity-centric KGQA.

Abstract

Knowledge Graph Question Answering (KGQA) has largely focused on entity-centric queries that return a single answer entity. However, many real-world questions are inherently relational, aiming to understand how entities are associated rather than which entity satisfies a query. In this work, we introduce relation-centric KGQA, a complementary setting in which the answer is a subgraph that represents the semantic relations among entities. The main challenge lies in the abundance of candidate subgraphs, where trivial or overly common connections often obscure the identification of unique and informative answers. To tackle this, we propose UniRel, a unified modular framework that combines a subgraph retriever with an LLM fine-tuned using reinforcement learning. The framework uses a reward function to prefer compact and specific subgraphs with informative relations and low-degree intermediate entities. Experiments show that UniRel improves connectivity and reward over Prompting baselines and generalizes well to unseen entities and relations. Moreover, UniRel can be applied to conventional entity-centric KGQA, achieving competitive or improved performance in several settings.