Thank you for the knowledge!

In 1999, the Open Mind Common Sense project began doing something unprecedented: it collected common sense knowledge by simply asking people to contribute it over the Web.

The knowledge was originally very unstructured. You could type sentences into a simple box, with the intent that we would later discover patterns in what people typed to turn the sentences into extracted knowledge.

Over time we developed various interfaces to knowledge collection to make the collected statements more informative to a computer, refine the existing knowledge, and fill in gaps in the knowledge base. We also created ConceptNet, a semantic graph of all the parseable knowledge in OMCS. We began collaborating with groups around the world, creating interfaces and games with a purpose to collect knowledge in many languages.

As a result, we were able to build ConceptNet, a large, multilingual semantic network that can help computers understand what text really means.

Crowdsourced knowledge has gotten bigger.

There are lots of projects collecting crowdsourced knowledge now, and they've got better systems and better interfaces than a small research group can build. ConceptNet is now actively expanding its breadth of knowledge, by pulling in other free knowledge sources besides OMCS.

We particularly admire Wikipedia and Wiktionary for the way they have built large communities, providing knowledge to the world and organizing some of it in a form that computers can parse.

ConceptNet 5 gets its knowledge from the existing data in Open Mind Common Sense, the various related Commonsense Computing projects, from various Games with a Purpose that we have collaborated with, from curated resources such as WordNet and JMDict, and from Wikipedia (via DBPedia) and Wiktionary.

How do I contribute now?