MDER-DR: Multi-Hop Question Answering with Entity-Centric Summaries

TL;DR

This paper introduces MDER-DR, a novel KG-based QA framework that enhances multi-hop question answering by generating entity-centric summaries and decomposing queries, significantly improving retrieval performance over traditional RAG methods.

Contribution

The paper presents MDER-DR, a new indexing and retrieval approach that improves multi-hop QA over knowledge graphs by avoiding explicit graph traversal and enabling better reasoning.

Findings

Achieves up to 66% improvement over standard RAG baselines.

Maintains robustness across multiple languages.

Effectively handles sparse and incomplete relational data.

Abstract

Retrieval-Augmented Generation (RAG) over Knowledge Graphs (KGs) suffers from the fact that indexing approaches may lose important contextual nuance when text is reduced to triples, thereby degrading performance in downstream Question-Answering (QA) tasks, particularly for multi-hop QA, which requires composing answers from multiple entities, facts, or relations. We propose a domain-agnostic, KG-based QA framework that covers both the indexing and retrieval/inference phases. A new indexing approach called Map-Disambiguate-Enrich-Reduce (MDER) generates context-derived triple descriptions and subsequently integrates them with entity-level summaries, thus avoiding the need for explicit traversal of edges in the graph during the QA retrieval phase. Complementing this, we introduce Decompose-Resolve (DR), a retrieval mechanism that decomposes user queries into resolvable triples and grounds them in the KG via iterative reasoning. Together, MDER and DR form an LLM-driven QA pipeline that is robust to sparse, incomplete, and complex relational data. Experiments show that on standard and domain specific benchmarks, MDER-DR achieves substantial improvements over standard RAG baselines (up to 66%), while maintaining cross-lingual robustness. Our code is available at https://github.com/DataSciencePolimi/MDER-DR_RAG.