NGDB-Zoo: Towards Efficient and Scalable Neural Graph Databases Training
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917293861109760 |
|---|---|
| author | Xie, Zhongwei Bai, Jiaxin Liu, Shujie Huang, Haoyu Li, Yufei Gao, Yisen Tsang, Hong Ting Song, Yangqiu |
| author_facet | Xie, Zhongwei Bai, Jiaxin Liu, Shujie Huang, Haoyu Li, Yufei Gao, Yisen Tsang, Hong Ting Song, Yangqiu |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_21597 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | NGDB-Zoo: Towards Efficient and Scalable Neural Graph Databases Training Xie, Zhongwei Bai, Jiaxin Liu, Shujie Huang, Haoyu Li, Yufei Gao, Yisen Tsang, Hong Ting Song, Yangqiu Machine Learning 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. |
| title | NGDB-Zoo: Towards Efficient and Scalable Neural Graph Databases Training |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.21597 |