Our Overload in 5 Pictures:
This is the server we had. It was very good at serving data in our test set for 20 languages.
Then we added more data. More than 1.2 million records in two dozen languages, with around 10 million joints.
With each record served as a full web page, and every search engine on the planet crawling all our pathways so the millions of new multilingual translations could be discovered, our little server could not handle the Big Data load.
This is what we need: a much stronger server, with a much more efficient way of interconnecting data.
And this is what we need to get there.
Please help get us back on the road by testing your favorite button below - they work best with a long string of zeros!
Residents of the United States must select the top button.
Residents of any country except the United States who wish to join or donate to the Kamusi Project, including US citizens using an address outside of the United States, should do so by choosing your favorite currency from those listed below.
The Kamusi Project is a participatory international effort dedicated to improving knowledge of the world's languages. Our long term mission is to produce dictionaries and other language resources for every language, and to make those resources available for free to everyone. To our enormous frustration, our search features are currently offline while we resolve interrelated server and financial issues.
Our secure donations page - the best way to give to Kamusi, your generosity will help us get back to full power right away!
Kamusi is different: We are producing dictionaries for each language, with full monolingual definitions for every concept. Each concept is then matched to similar ideas in all other participating languages. Watch a video of the system in action - a dictionary really can do that! We are constantly adding new terms and new languages - contact us if you can contribute for your language.
Big Data: We have grown to well over a million interlinked terms in more and more languages! Much of this is "seed" data that does not yet meet our quality goals - watch this space for news about the games and tools we will soon introduce for you to help improve and increase knowledge about all the world's words.
Take the Kamusi Challenge: We contend that Kamusi is the world's most effective model for translating terms among a large number of languages, and we invite you to test that proposition on our demonstration data. See more...
Kamusi is a work in progress, so expect rough edges as we race to improve our features and our data. January 2016: We apologize for any server outages you might experience. Serving well over 10 million new links, our existing hardware has been overtaxed. We are working to resolve the problems. Meanwhile, unfortunately, SEARCH IS OFFLINE. This project has emerged from the Internet Living Swahili dictionary (we need your help adding English and Swahili definitions!), and has expanded to many additional African languages. Now we are incorporating languages from Europe, Asia, Australia, and the Americas - please join us, help us grow, and come back often!
1.2 Million New Records. Good, but not good enough. Now you can find your words with confidence...
People often ask, "What is the difference is between Kamusi and other projects that deal with...
Molecular Lexicography: A Lexical Data Model for Human Language Technology Introduction: The Kamusi...