Rob Speer edited this page Nov 4, 2016 · 55 revisions

ConceptNet 5

ConceptNet is a multilingual knowledge base, representing words and phrases that people use and the common-sense relationships between them. The knowledge in ConceptNet is collected from a variety of resources, including crowd-sourced resources (such as Wiktionary and Open Mind Common Sense), games with a purpose (such as Verbosity and nadya.jp), and expert-created resources (such as WordNet and JMDict).

You can browse what ConceptNet knows at http://conceptnet.io.

General information about ConceptNet

  • FAQ
  • Changelog
  • Web API - A quick way to get started with ConceptNet 5, this REST API lets you query the knowledge about any node in ConceptNet, and even make some inferences about which nodes are related.

Running your own copy

  • Running your own copy - How to run your own copy of ConceptNet, or rebuild it with your own modifications, using Docker to manage the process.
  • Build process - How to build the ConceptNet data locally on a Unix system (even without Docker), and what you get as a result.
  • Copying and sharing ConceptNet - ConceptNet is open data, and we want you to copy it and put it to new uses.
  • Downloads

Details about the data

  • Languages - The hundreds of languages (and language families) represented in ConceptNet.
  • Edges - The basic units of knowledge in ConceptNet and what they represent.
  • Relations - Each edge has a relation. These are the kinds of relations we're using so far, and what they mean.
  • URI hierarchy - What a URI such as /s/contributor/omcs/dev means.

Getting help

If you have a bug report or feature request for ConceptNet, please post it on our GitHub issues page.

To get help using ConceptNet, and to get announcements of upgrades and API downtime, you should join the conceptnet-users mailing list.

ConceptNet has a chat room on Gitter: https://gitter.im/commonsense/conceptnet5