提升神经图数据库训练效率，支持大规模复杂推理
NGDB-Zoo: Towards Efficient and Scalable Neural Graph Databases Training
- 将逻辑操作与查询结构解耦，实现动态数据流并行训练
- 在六大数据集上提速1.8至6.8倍，保持高GPU利用率
- 融合预训练文本模型语义先验，避免内存溢出和I/O阻塞
神经图数据库（NGDB）能够在不完整知识结构上进行复杂逻辑推理，但其训练效率与表达能力受限于固定的查询级批处理和仅依赖结构的嵌入方式。本文提出NGDB-Zoo，通过操作级训练与语义增强的协同设计，突破上述瓶颈。该框架将逻辑操作从查询拓扑中解耦，将训练过程转化为动态调度的数据流执行，实现多流并行，相较基线提升1.8×至6.8×吞吐量。此外，我们构建解耦架构，无缝集成预训练文本编码器（PTEs）提供的高维语义先验，避免引入I/O阻塞或内存溢出。在六个基准数据集（包括ogbl-wikikg2和ATLAS-Wiki等大规模图）上的大量实验表明，NGDB-Zoo可在多种逻辑模式下维持高效GPU利用率，并显著降低混合神经符号推理中的表征摩擦。
原文摘要 · Abstract (English)
Neural Graph Databases (NGDBs) facilitate complex logical reasoning over incomplete knowledge structures, yet their training efficiency and expressivity are constrained by rigid query-level batching and structure-exclusive embeddings. We present NGDB-Zoo, a unified framework that resolves these bottlenecks by synergizing operator-level training with semantic augmentation. By decoupling logical operators from query topologies, NGDB-Zoo transforms the training loop into a dynamically scheduled data-flow execution, enabling multi-stream parallelism and achieving a $1.8\times$ - $6.8\times$ throughput compared to baselines. Furthermore, we formalize a decoupled architecture to integrate high-dimensional semantic priors from Pre-trained Text Encoders (PTEs) without triggering I/O stalls or memory overflows. Extensive evaluations on six benchmarks, including massive graphs like ogbl-wikikg2 and ATLAS-Wiki, demonstrate that NGDB-Zoo maintains high GPU utilization across diverse logical patterns and significantly mitigates representation friction in hybrid neuro-symbolic reasoning.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。