查询感知的扩散激活用于知识图谱上的多跳检索
Query-Aware Spreading Activation for Multi-Hop Retrieval over Knowledge Graphs
AI总结：
提出一种查询感知的扩散激活方法，通过单步语义门实现查询感知遍历，并用单一Cypher查询完成整个检索过程，在MuSiQue上匹配QAFD-RAG并超越HippoRAG，同时降低延迟。
摘要：
基于知识图谱的检索增强生成（Graph RAG）通过利用图结构，在多跳问答中优于平面段落检索。然而，在大多数现有系统中，问题仅设置种子节点；后续遍历变得“查询盲目”，仅依赖于图结构。例外是QAFD-RAG，它通过结合边重新加权的流扩散求解器实现查询感知遍历。这种架构需要将完整图加载到Python内存中，并使用迭代求解器，迭代次数可变，增加了与图数据库集成的复杂性。我们提出一种扩散激活方法，通过单步语义门实现相同的查询感知遍历：步骤权重是候选实体描述与问题之间的余弦相似度，迭代次数固定。整个检索过程——种子映射、传播、top-K选择和上下文组装——表示为单个Cypher查询，在Neo4j的一次往返中执行；图从不离开数据库。在MuSiQue上，我们的方法在精确匹配上匹配QAFD-RAG（32.80 vs 33.50），并在我们的比较中最强的纯结构基线HippoRAG上超出5.3 EM和3.4 F1；在2WikiMultiHopQA上，HippoRAG和QAFD-RAG由于其短语节点架构保持优势。禁用门的消融实验证实，门是同时获得3.6至7.4个F1点增益和1.5至4.9倍检索延迟降低的来源。
英文摘要：
Retrieval-augmented generation built on knowledge graphs (Graph RAG) outperforms flat passage retrieval on multi-hop question answering by leveraging graph structure. In most existing systems, however, the question only sets the seed nodes; the subsequent traversal becomes "query-blind", depending solely on the graph structure. The exception is QAFD-RAG, which implements query-aware traversal via a flow-diffusion solver with combined edge re-weighting. This architecture requires loading the full graph into Python memory and an iterative solver with a variable number of iterations complicating integration with the graph database. We propose a spreading-activation method that achieves the same query-aware traversal with a single per-step semantic gate: the step weight is the cosine similarity between the candidate entity's description and the question, and the number of iterations is fixed. The whole retrieval procedure - seed mapping, propagation, top-K selection and context assembly - is expressed as a single Cypher query executed in one round-trip to Neo4j; the graph never leaves the database. On MuSiQue our method matches QAFD-RAG by exact match (32.80 vs 33.50) and outperforms the strongest purely-structural baseline in our comparison, HippoRAG, by 5.3 EM and 3.4 F1; on 2WikiMultiHopQA HippoRAG and QAFD-RAG retain an advantage due to their phrase-node architectures. An ablation with the gate disabled confirms that the gate is the source of a simultaneous F1 gain of 3.6 to 7.4 points and a retrieval-latency reduction by a factor of 1.5 to 4.9.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
