英伟达开源Auto Ontology语义上下文库
Start here
2026年11月3日，英伟达宣布基于Apache 2.0协议开源Auto Ontology库。该项目旨在利用企业元数据自动构建语义上下文层，以解决AI智能体在数据分析以及自然语言转SQL过程中出现的指标失真和定义不一致问题。Auto Ontology支持通过模型上下文协议（MCP）、命令行界面（CLI）和API为智能体提供经过校验的业务定义与数据血缘验证，同时默认集成了NVIDIA NIM推理端点，并支持容器化部署。
AI-generated from reports · Updated 1 day ago
Report timeline
Follow the reports to see different sides of the story.
- Buzzing HN · 中文精选SelectedNVIDIA Open-Sources Auto Ontology Semantic Context Library
NVIDIA has open-sourced the Auto Ontology library, designed to automatically construct a semantic context layer from enterprise metadata to address metric distortions and inconsistent definitions in AI agents during data analytics and text-to-SQL tasks. Released under the Apache 2.0 license, the project supports delivering verified business definitions and data lineage validation to agents via MCP, CLI, and APIs, with default integration of NVIDIA NIM inference endpoints and support for containerized deployment.
Story heat trend
Current heat 5·Comparable peak 10（Nov 3 22:00）·Comparable change, last 24 hours–
Trends compare only the same parties observed continuously, so the scope can be narrower than current heat counts. Move the pointer or tap the chart to see hourly heat; use the left and right arrow keys on a keyboard.