ConceptNet resource

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

2953 stars · 357 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

ConceptNet is a large multilingual commonsense knowledge graph linking over 13 million concepts across many natural languages, distributed under a Creative Commons license. This repository contains the Python toolset for building ConceptNet 5 from raw data, plus the HTML interface and JSON Web API for serving it.

Use cases

• find commonsense relationships between words and concepts
• get multilingual semantic knowledge for NLP models
• build a custom knowledge graph from my own data
• query a semantic network via a JSON API
• download pretrained word embeddings like Numberbatch
• use commonsense knowledge as input for machine learning on text

When to choose

• you need broad, multilingual commonsense knowledge for NLP or ML research
• you want to build a customized knowledge graph from your own data sources
• you need pretrained semantic embeddings (ConceptNet Numberbatch) for downstream models

When to avoid

• you only need to query ConceptNet - use the public Web API instead of building it yourself
• you lack the required resources (30 GB RAM, Unix environment) for the build process
• you need up-to-date, actively developed tooling - the project is in maintenance mode

Facets

dataset · maturity maintenance

nlp machine-learning search-engine api-framework etl artificial-intelligence machine-learning databases apis python cli knowledge-graph commonsense-knowledge semantic-network multilingual word-embeddings numberbatch natural-language-processing linux web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="commonsense/conceptnet5")