Sources and terms

Word vectors

Closeness between words is computed with the fastText vectors (cc.ko.300 and cc.en.300) released by Facebook AI Research. They are 300 dimensional vectors trained on web text and Wikipedia.

These vectors are distributed under the Creative Commons Attribution-Share-Alike License 3.0. The terms are to credit the source and share alike. Commercial use is allowed.

From those vectors we keep the 50,000 most common words and precompute each day's cosine to the answer into files. Those derived files are shared under the same terms (CC BY-SA 3.0).

Answer candidate lists

Korean candidates come from the Wiktionary appendix of 5,800 frequent Korean words. English candidates come from Wiktionary Frequency lists/PG/2006/04/1-10000.

Wiktionary text is provided under the Creative Commons Attribution-Share-Alike License. The filtering criteria are written in docs/filters.md.

Made by

Made by Gwiraedang (gwirae.com). No name, art, text or code was taken from other games. Only the general kind of game is shared.