Catchtype word data

These files are word lists and statistics that Catchtype (https://catchtype.com) uses to tell
Korean typed on an English keyboard layout from English. They are shared under the same license as
the list they were made from.

Source: FrequencyWords 2018 Korean and English 50k lists (OpenSubtitles 2018), by Hermit Dave,
https://github.com/hermitdave/FrequencyWords, licensed under CC BY-SA 4.0. FrequencyWords was counted
from the OpenSubtitles 2018 corpus (OPUS, http://opus.nlpl.eu/OpenSubtitles2018.php; P. Lison and
J. Tiedemann, 2016), made from subtitles at http://www.opensubtitles.org.

Changes: the lists were cut and filtered, and turned into the statistics below. Counts from NSMC
(https://github.com/e9t/nsmc) and Mozilla Common Voice sentences, both CC0 1.0, and from word lists
written for Catchtype were added to the syllable and letter statistics.

License: these files are licensed under the Creative Commons Attribution-ShareAlike 4.0
International License by Hebbian AI Co., Ltd. See LICENSE.txt or
https://creativecommons.org/licenses/by-sa/4.0/.

Files:
- korean-words.txt: common Korean words, most common first, one per line.
- korean-more-words.txt: the less common Korean words that follow them, most common first.
- english-words.txt: common English words, most common first, one per line.
- english-subtitle-only-words.txt: the words among the first 20,000 of english-words.txt that are not in
  Google Books Ngram 2020 (CC BY 3.0) or SCOWL 2020.12.07; those lists were only used to remove words.
- common-syllables.txt: common Hangul syllables, one per line.
- korean-syllable-pairs.txt: two Hangul syllables and a numerical value per line.
- korean-unseen.txt: a Hangul syllable and a numerical value per line.
- english-letter-triples.txt: three letters and a numerical value per line.
- english-unseen.txt: two letters and a numerical value per line.
- unknown.txt: a single numerical value.