The PanLex project is the major activity of Utilika Foundation, in collaboration with the University of Washington Turing Center. You can read about it in the description of the foundation’s work and at the project’s technical collaboration site. In one sentence: PanLex aims to be a translating dictionary containing all the words of all the languages in the world.
In developing PanLex we rely on published and live sources. The principal initial sources are about three thousand bi- and multilingual dictionaries.
The first demonstrations of the applications that PanLex makes possible were the Turing Center’s multilingual image search engine, PanImages, and its Panlingual Translator.
A prototype of a self-localizing PanLex user interface is available for you to test. [WARNING: Safari 5.1 is extremely slow processing this interface. That bug has been reported.] Begin by choosing the language you want the interface to communicate in. At this time, some languages are more complete and work faster than others. Those that work fastest are English, Esperanto, French, German, Norwegian, Russian, and Turkish.
Work began in 2009 on a prototype of a PanLex application programming interface (API). At present, there is a primitive test client for this API in English, Esperanto, and Turkish. The client translates words from the interface language into other languages.
Those working on PanLex have created some developer documentation.
