发表机构
Heinz Nixdorf Institute, Paderborn University（海因茨·尼克斯多夫研究所，帕德博恩大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对大型语言模型在知识密集型任务中可靠性受限的问题，提出MARS方法，通过结构化检索过程链接问题实体与知识图谱，迭代获取信息并决定检索深度或生成SPARQL查询，在多个基准测试中性能有竞争力且高效可扩展。
AI 中文摘要
大型语言模型（LLMs）在推理性能上表现强劲，但产生幻觉的倾向限制了其在需要最新且可靠信息的知识密集型任务中的可靠性。将知识图谱（KGs）与LLMs相结合，有助于利用可不断更新且无需高成本微调的显式符号知识，同时受益于快速发展的LLM推理。我们提出了MARS，一种无需模型微调的可扩展知识图谱问答（KGQA）方法。MARS执行结构化检索过程，将问题实体链接到KG并迭代检索相关的下一跳信息，而非依赖开放式智能体探索。在每一步，MARS决定是继续图遍历还是生成最终的SPARQL查询，使模型能根据问题调整检索深度。我们在多个LLMs和设置下的三个既定KGQA基准上评估了MARS，包括多语言评估，并通过消融研究和错误分析提供见解。我们的方法相对于现有方法具有有竞争力的性能，同时保持高效和可扩展性。评估结果、代码和资源可公开获取。
英文摘要
Large language models (LLMs) have demonstrated strong reasoning performance, but their tendency to hallucinate limits their reliability in knowledge-intensive tasks requiring up-to-date and grounded information. Combining knowledge graphs (KGs) with LLMs facilitates the use of explicit symbolic knowledge that can be continuously updated without costly fine-tuning, while benefiting from rapidly advancing LLM reasoning. We propose MARS, a scalable knowledge graph question answering (KGQA) approach that requires no model fine-tuning. Rather than relying on open-ended agentic exploration, MARS performs a structured retrieval procedure that links question entities to the KG and iteratively retrieves relevant next-hop information. At each step, MARS decides whether to continue graph traversal or to generate the final SPARQL query, allowing the model to adapt the retrieval depth to the question while keeping the overall pipeline more predictable than fully agentic approaches. We evaluate MARS on three established KGQA benchmarks across several LLMs and settings, including multilingual evaluation, and provide insights through ablation studies and error analysis. Our approach achieves competitive performance relative to state-of-the-art methods while remaining efficient and scalable. The evaluation results, code and resources are publicly available: https://github.com/dice-group/mars-kgqa.
CommentsEKAW 2026 (accepted-posters-and-demos" target="_blank" rel="noopener">https://ekaw2026.di.unito.it/accepted-posters-and-demos)
