能力标签
🛠
AI工具

本体知识图谱编辑器

基于 TypeScript · 开源免费,本地部署,数据完全自主可控
英文名:ontosphere
⭐ 135 Stars 🍴 18 Forks 💻 TypeScript 📄 NOASSERTION 🏷 AI 7.8分
7.8AI 综合评分
知识图谱RDF编辑语义网JSON-LD图编辑器
✦ AI Skill Hub 推荐

AI Skill Hub 推荐使用:本体知识图谱编辑器 是一款优质的AI工具。AI 综合评分 7.8 分,在同类工具中表现稳健。如果你正在寻找可靠的AI工具解决方案,这是一个值得深入了解的选择。

📚 深度解析

本体知识图谱编辑器 是一款基于 TypeScript 的开源工具,在 GitHub 上收获 0k+ Star,是知识图谱、RDF编辑、语义网、JSON-LD领域中的优质开源项目。开源工具的最大优势在于代码完全透明,你可以审计每一行代码的安全性,也可以根据自身需求进行二次开发和定制。

**为什么要使用开源工具而非商业 SaaS?**
对于个人开发者和有隐私需求的用户,本地部署的开源工具意味着数据不离本机,不受第三方服务商的数据政策约束。同时,开源工具通常没有使用次数限制和月度费用,一次安装即可长期使用,对于高频使用场景的总拥有成本(TCO)远低于订阅制商业工具。

**安装与环境准备**
本体知识图谱编辑器 依赖 TypeScript 运行环境。建议通过 pyenv(Python)或 nvm(Node.js)管理 TypeScript 版本,避免全局环境污染。对于新手用户,推荐先创建虚拟环境(python -m venv venv && source venv/bin/activate),再安装依赖,这样即使出现问题也可以随时删除虚拟环境重新开始,不影响系统稳定性。

**社区与维护**
GitHub Issue 和 Discussion 是获取帮助的最快渠道。在提问前建议先检查 Closed Issues(已关闭的问题),大多数常见问题都已有解答。遇到 Bug 时,提供 pip list 的输出、完整错误堆栈和最小可复现示例,能显著提高开发者响应速度。AI Skill Hub 将持续追踪 本体知识图谱编辑器 的版本更新,及时通知重要功能变化。

📋 工具概览

基于浏览器的RDF/本体知识图谱可视化编辑工具。支持从文件、URL加载RDF数据,提供图形化编辑界面。适合语义网、链接数据、知识图谱研究者和开发者使用。

本体知识图谱编辑器 是一款基于 TypeScript 开发的开源工具,专注于 知识图谱、RDF编辑、语义网 等核心功能。作为 GitHub 开源项目,它拥有活跃的社区支持和持续的版本迭代,代码完全透明可审计,支持本地部署以保护数据隐私。无论是个人使用还是集成到企业工作流,都能提供稳定可靠的解决方案。

GitHub Stars
⭐ 135
开发语言
TypeScript
支持平台
Windows / macOS / Linux
维护状态
轻量级项目,按需更新
开源协议
NOASSERTION
AI 综合评分
7.8 分
工具类型
AI工具
Forks
18

📖 中文文档

以下内容由 AI Skill Hub 根据项目信息自动整理,如需查看完整原始文档请访问底部「原始来源」。

基于浏览器的RDF/本体知识图谱可视化编辑工具。支持从文件、URL加载RDF数据,提供图形化编辑界面。适合语义网、链接数据、知识图谱研究者和开发者使用。

本体知识图谱编辑器 是一款基于 TypeScript 开发的开源工具,专注于 知识图谱、RDF编辑、语义网 等核心功能。作为 GitHub 开源项目,它拥有活跃的社区支持和持续的版本迭代,代码完全透明可审计,支持本地部署以保护数据隐私。无论是个人使用还是集成到企业工作流,都能提供稳定可靠的解决方案。

📌 核心特色
  • 开源免费,支持本地部署,数据完全自主可控
  • 活跃的 GitHub 开源社区,持续迭代更新
  • 提供详细文档和使用示例,新手友好
  • 支持自定义配置,灵活适配不同使用环境
  • 可作为基础组件集成进现有技术栈或进行二次开发
🎯 主要使用场景
  • 本地部署运行,保护数据隐私,满足合规要求
  • 自定义集成到现有系统,扩展技术栈能力
  • 作为开源基础组件进行商业化二次开发
以下安装命令基于项目开发语言和类型自动生成,实际以官方 README 为准。
安装命令
# 方式一:npm 全局安装
npm install -g ontosphere

# 方式二:npx 直接运行(无需安装)
npx ontosphere --help

# 方式三:项目依赖安装
npm install ontosphere

# 方式四:从源码运行
git clone https://github.com/ThHanke/ontosphere
cd ontosphere
npm install
npm start
📋 安装步骤说明
  1. 访问 GitHub 仓库页面
  2. 按照 README 文档完成依赖安装
  3. 根据系统环境完成初始化配置
  4. 参考官方示例或文档开始使用
  5. 遇到问题可在 GitHub Issues 中查找解答
以下用法示例由 AI Skill Hub 整理,涵盖最常见的使用场景。
常用命令 / 代码示例
# 命令行使用
ontosphere --help

# 基本用法
ontosphere [options] <input>

# Node.js 代码中使用
const ontosphere = require('ontosphere');

const result = await ontosphere.run(options);
console.log(result);
以下配置示例基于典型使用场景生成,具体参数请参照官方文档调整。
配置示例
# ontosphere 配置说明
# 查看配置选项
ontosphere --config-example > config.yml

# 常见配置项
# output_dir: ./output
# log_level: info
# workers: 4

# 环境变量(覆盖配置文件)
export ONTOSPHERE_CONFIG="/path/to/config.yml"
📑 README 深度解析 真实文档 完整度 77/100 查看 GitHub 原文 →
以下内容由系统直接从 GitHub README 解析整理,保留代码块、表格与列表结构。

Ontosphere — Browser-based RDF Knowledge Graph Editor

DOI License

I want to…Start here
Try the live demo[Open Ontosphere ↗](https://thhanke.github.io/ontosphere)
Connect an AI agent[AI / MCP Integration](#ai--mcp-integration)
Run it locally[Quick start (development)](#quick-start-development)
Load my own data[Startup / URL parameters](#startup--url-parameters)
Contribute code[Contributing](#contributing--development-notes)

Overview

Ontosphere is a browser-based RDF/ontology knowledge graph editor. It loads RDF from local files, remote URLs, or SPARQL/Fuseki endpoints; lets users author nodes and edges directly on the canvas; runs OWL 2 DL reasoning (via Konclude) with visual differentiation of inferred triples; and applies multi-algorithm layout (Dagre, ELK) and automatic clustering for large graphs. Additional features include namespace management with live URI renaming, a drag-and-drop workflow template catalog, and a Model Context Protocol (MCP) server for AI-agent integration. All computation runs entirely client-side in the browser against an in-memory RDF store backed by Web Workers — no backend required.

Key capabilities

  • Load RDF/Turtle/JSON-LD/RDF-XML/N-Triples from local files or remote URLs (including SPARQL endpoints and Fuseki datasets).
  • Startup URL support: auto-load an RDF file via URL query parameter (see "Startup / URL usage" below).
  • Reactodia canvas: pan, zoom, minimap, fit-view, with entity group (cluster) support and smooth animations.
  • Authoring mode (always on): add nodes via search, draw edges by dragging the halo "Establish Link" handle, edit node annotation properties and link predicates directly on the canvas. Undo/Redo support. Entity auto-complete uses scored domain/range tiers derived from loaded ontologies.
  • Search: type in the search box to find entities by label or IRI; press Enter to cycle through matches on the canvas.
  • TBox / ABox views: toggle between ontology-level classes/properties (TBox) and data-level individuals (ABox).
  • Layout engine: multiple algorithms — Dagre (horizontal/vertical), ELK (layered, force, stress, radial), and Reactodia-default — all running in Web Workers so the UI stays responsive. Spacing is adjustable via a slider; re-layout triggers automatically when spacing changes.
  • Clustering: automatic grouping of large graphs on load. Three algorithms available — Label Propagation (default), Louvain, and K-Means. Threshold is configurable (default 100 nodes). Expand/collapse individual clusters or all at once from the toolbar.
  • DL reasoning (Konclude): run OWL 2 DL inference in the browser and see inferred triples rendered as amber dashed edges; inferred types/annotations appear in amber italic. A reasoning report lists all inferred triples. Includes automatic OWL DL consistency checking — the Errors tab shows per-entity clash details when the ontology is contradictory. Clear inferred triples any time without affecting asserted data.
  • Namespace management: edit namespace URIs directly in the legend panel (rename propagates across all stored triples). Colour-coded namespace badges on nodes and edges.
  • Export the current graph as Turtle, RDF/XML, or JSON-LD.
  • Workflow catalog: drag reusable workflow template cards from the sidebar onto the canvas to instantiate connected subgraphs.
  • MCP support: exposes a Model Context Protocol server (via the browser's navigator.modelContext API) for AI-agent integration. Tools: loadRdf, loadOntology, suggestOntologiesForTask, queryGraph, exportGraph, exportImage, addNode, removeNode, expandNode, getNodes, addLink, removeLink, getLinks, runLayout, clusterNodes, layoutNodes, focusNode, fitCanvas, runReasoning, clearInferred, getNeighbors, findPath, getNodeDetails, updateNode, getGraphState, setNamespace, removeNamespace, listNamespaces, loadShacl, validateGraph, getCapabilities, help. MCP manifest at /.well-known/mcp.json.

Setup (Playwright / headless)

navigator.modelContext does not exist in headless Chromium. Inject the polyfill before the page loads using page.addInitScript:

await page.addInitScript(() => {
  const tools = {};
  Object.defineProperty(navigator, 'modelContext', {
    value: { registerTool: async (n, _d, _s, h) => { tools[n] = h; } },
    configurable: true,
  });
  window.__mcpTools = tools;
});

// After page load:
await page.evaluate(async () => {
  const mod = await import('/src/mcp/ontosphereMcpServer.ts');
  await mod.registerMcpTools();
});

// Call a tool:
await page.evaluate(async ([name, params]) => window.__mcpTools[name](params),
  ['addNode', { iri: 'ex:alice', typeIri: 'foaf:Person', label: 'Alice' }]);

In a browser with native navigator.modelContext, tools register automatically on app load.

Quick start (development)

1. Install dependencies:

   npm install
   
2. Start the Vite dev server:
   npm run dev
   
3. Open in your browser:
   http://localhost:8080/
   

Full example (CKAN private dataset via Fuseki SPARQL)

http://docker-dev.iwm.fraunhofer.de:8080/
  ?rdfUrl=https://docker-dev.iwm.fraunhofer.de/dataset/<uuid>/fuseki/$/sparql
  &apiKey=<ckan-api-jwt-token>

Example output

An AI agent built this from scratch in one session — full demo with tool calls →

FOAF social network

Reasoning demo

The reasoning demo showcases OWL 2 DL / SROIQ(D) inference on a small employee ontology: Open demo ↗

The demo (public/reasoning-demo.ttl) defines a Person → Employee → Manager → Executive hierarchy with ABox assertions that drive inference patterns across all OWL 2 DL construct groups:

OWL 1 RL patterns: 1. rdfs:subPropertyOfex:hasFriend sub-property of ex:knows: alice hasFriend bobalice knows bob. 2. owl:inverseOfex:isManagedBy inverse of ex:manages: alice manages carolcarol isManagedBy alice. 3. owl:SymmetricPropertyex:isColleagueOf is symmetric: bob isColleagueOf carol → reverse direction. 4. owl:TransitivePropertyex:hasSupervisor is transitive: bob→alice, alice→davebob→dave. 5. rdfs:domainex:dave has no type; because he is subject of ex:manages (domain ex:Manager), the reasoner infers dave rdf:type ex:Manager.

OWL 2 DL extensions: 6. owl:someValuesFromalice and carol each worksOn projectAlpha (a Project) → inferred ProjectContributor. 7. owl:hasValuecarol isManagedBy alice (via inverseOf) → carol inferred DirectReport (hasValue restriction on alice). 8. owl:intersectionOfdave manages bob (inferred Manager) and eve (Employee) → dave inferred TeamLead. 9. owl:disjointWithContractor disjointWith Employee; frank is a Contractor (structural TBox constraint). 10. owl:complementOfNonEmployee ≡ ¬Employee (structural TBox only). 11. owl:propertyChainAxiomhasGrandManager ← hasSupervisor ∘ hasSupervisor: carol→bob→alicecarol hasGrandManager alice. 12. owl:unionOfLeadershipTeam ≡ Executive ∪ Manager: alice (Executive) and dave (inferred Manager) → inferred LeadershipTeam. 13. owl:sameAsaliceCEO sameAs alice: aliceCEO inherits all of alice's inferred types including Executive.

A separate inconsistency demo (public/reasoning-demo-inconsistent.ttl) shows the consistency checker in action: Open inconsistency demo ↗

inc:frank is asserted as both inc:Employee and inc:Contractor, which are declared owl:disjointWith. Running reasoning produces isConsistent: false, reasoning is skipped, and the report's Errors tab shows the disjointness clash on frank.

Demo

DemoFinal state
**[FOAF social network](docs/mcp-demo/foaf-social-network.md)**<br>Build a social network, extend FOAF with employment classes, run reasoning[![FOAF social network final state](docs/mcp-demo/foaf-social-network/04-frank-focus.svg)](docs/mcp-demo/foaf-social-network.md)
**[DL reasoning (Konclude)](docs/mcp-demo/reasoning-demo.md)**<br>Build TBox + ABox, infer types via domain/range and transitivity[![DL reasoning final state](docs/mcp-demo/reasoning-demo/04-dave-focus.svg)](docs/mcp-demo/reasoning-demo.md)
**[Scene ontology](docs/mcp-demo/scene-ontology.md)**<br>Load an external ontology, author individuals, export Turtle[![Scene ontology final state](docs/mcp-demo/scene-ontology/04-jake-focus.svg)](docs/mcp-demo/scene-ontology.md)
**[Manchester Pizza Tutorial](docs/mcp-demo/pizza-tutorial.md)**<br>Full OWL pizza ontology — classes, disjointness, properties, DL reasoning[![Manchester Pizza Tutorial final state](docs/mcp-demo/pizza-tutorial/20-owa-vegetarian-lesson.svg)](docs/mcp-demo/pizza-tutorial.md)

Regenerate:

```sh npm run demo:all

Recording demo videos

See docs/demo-scripts/HOWTO.md for the full guide.

Two styles of demo video are supported:

Seed-driven — write a seed markdown file in docs/mcp-demo/seeds/ with JSON-RPC tool calls embedded in backtick blocks. The runner parses the seed and executes each turn directly against window.__mcpTools. Existing seeds: foaf-social-network, reasoning-demo, scene-ontology, pizza-tutorial.

Chat-style (side-by-side) — open demo-stage.html (mock chat left, app right), inject messages programmatically via addChatMessage(), and call tools on the app iframe via callToolOnStage(). No relay popup needed. Example: pizza-tutorial-chat.

Demo videos

VideoDescription
[advert-intro.mp4](docs/demo-videos/advert-intro.mp4)Relay bookmarklet — mock chat + Ontosphere side by side
[foaf-social-network.mp4](docs/demo-videos/foaf-social-network.mp4)AI builds a FOAF social graph with DL reasoning
[reasoning-demo.mp4](docs/demo-videos/reasoning-demo.mp4)AI builds an OWL ontology and runs DL reasoning (Konclude)
[scene-ontology.mp4](docs/demo-videos/scene-ontology.mp4)AI builds a film scene ontology on BFO/RO upper ontology
[pizza-tutorial.mp4](docs/demo-videos/pizza-tutorial.mp4)Manchester Pizza Ontology — class hierarchy, disjointness, DL reasoning
[pizza-tutorial-chat.mp4](docs/demo-videos/pizza-tutorial-chat.mp4)OWL pizza tutorial as AI tutor lesson, side-by-side chat

To re-record all videos:

npm run demo:video   # starts dev server, records, encodes, kills server

Authentication (API key)

ParameterDefaultDescription
apiKeyValue sent as an authentication header with the RDF fetch.
apiKeyHeaderAuthorizationName of the HTTP header.
?rdfUrl=https://private-endpoint.example.org/data.ttl
&apiKey=Bearer+my-token
&apiKeyHeader=Authorization

The API key is sent only with the RDF fetch request. CORS: the server must allow the Ontosphere origin with credentials (wildcard * origins are incompatible with authenticated requests).

AI / MCP Integration

Ontosphere exposes a full Model Context Protocol tool surface so AI agents can build and reason over knowledge graphs through natural-language chat.

Troubleshooting

  • rdfUrl doesn't load on open:
  • Confirm the URL is percent-encoded in the address bar.
  • Open DevTools → Network and check the fetch request and response headers.
  • Look for CORS errors (Access-Control-Allow-Origin).
  • Check the console for RDF parser errors or application diagnostics.
  • 403 when using certain query parameter names:
  • Some servers intercept reserved query names. Use ?rdfUrl=... to avoid conflicts.
  • Graph is very large / slow:
  • Increase the large-graph threshold in Settings or reduce the number of loaded triples.
  • Clustering activates automatically above the threshold; use Expand All sparingly on huge graphs.
🎯 aiskill88 AI 点评 A 级 2026-06-10

专业的RDF编辑工具,MCP集成设计合理。代码质量高,社区活跃度中等,适合语义网领域专业用户。

⚡ 核心功能

👥 适合人群

AI 技术爱好者研究人员和学生开发者和工程师技术创业者

🎯 使用场景

  • 本地部署运行,保护数据隐私,满足合规要求
  • 自定义集成到现有系统,扩展技术栈能力
  • 作为开源基础组件进行商业化二次开发

⚖️ 优点与不足

✅ 优点
  • +完全开源免费,无授权费用
  • +本地部署,数据完全自主可控
  • +开发者社区支持,遇问题可查可问
⚠️ 不足
  • 安装和初始配置可能需要一定技术基础
  • 功能完整性通常不如成熟商业产品
  • 技术支持主要依赖开源社区,响应速度不稳定
⚠️ 使用须知

该工具使用 NOASSERTION 协议,商用场景请仔细阅读协议条款,必要时咨询法律意见。

AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。

建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。

📄 License 说明

📄 NOASSERTION — 请查阅原始协议条款了解具体使用限制。

🔗 相关工具推荐

🧩 你可能还需要
基于当前 Skill 的能力图谱,自动补全的工具组合

❓ 常见问题 FAQ

支持RDF/XML、Turtle、N-Triples等标准格式,以及JSON-LD
💡 AI Skill Hub 点评

总体来看,本体知识图谱编辑器 是一款质量良好的AI工具,在同类工具中具备一定竞争力。AI Skill Hub 将持续追踪其更新动态,建议收藏备用,结合自身场景选择合适时机引入使用。

📚 深入学习 本体知识图谱编辑器
查看分步骤安装教程和完整使用指南,快速上手这款工具
🌐 原始信息
原始名称 ontosphere
Topics 知识图谱RDF编辑语义网JSON-LD图编辑器
GitHub https://github.com/ThHanke/ontosphere
License NOASSERTION
语言 TypeScript
🔗 原始来源
🐙 GitHub 仓库  https://github.com/ThHanke/ontosphere 🌐 官方网站  https://thhanke.github.io/ontosphere/

收录时间:2026-06-10 · 更新时间:2026-06-10 · License:NOASSERTION · AI Skill Hub 不对第三方内容的准确性作法律背书。